Mantis ASM Electronic Feeder - the thing that almost broke me

[Zdenko Stanec]

A coffee read ☕,

Hi there,

After a long year in the making, the thing that almost broke me is finally about to be frozen.

Please, don’t ask why I did it (obviously for the people in need 😅), or how much time and money it took to develop something like this, including choosing the right partners to manufacture all the custom parts integrated into the design.🤐 Was it worth it? Well, depends on what plans you have, for me... yes.

The purpose of this post is to gather insights and maybe help guide someone walking the same path.

The hardest part? Designing a system that’s scalable in width without requiring a bunch of new components keeping the BOM as tight as possible while maintaining the same mechanical principles across all sizes, basically designing all feeder widths at the same time in the head. The goal was to make every part affordable, easy to replace, user-friendly, long-lasting, and of course, as precise as possible.

P.S.
I’m not using N20 or N30 motors in this design, went with something more robust, more about that in the notes below. ☕

__________

If you're interested, please leave a comment, it’ll help me decide whether to make this available to others or keep it exclusive to my machine.

__________

C Sample tests, motors working on 10V instead of 12V, and also using only "Sprocket encoder" for positioning:

https://youtu.be/rHOv3Ly120w

__________

Say hello to Mantis,

Mantis SMT Feeder System: A Precise and Reliable Feeder for OpenPnP

The Mantis SMT Feeder is an electronic feeder system designed for the OpenPnP community. It focuses on precise and repeatable positioning, ease of use, and cost-effective maintenance. The Mantis Feeder provides reliable performance with a user-friendly interface.

Key Features and Design

The Mantis system consists of three main components:

• Mantis Feeder Bank [MFB] – Houses integrated EEPROMs with dedicated addresses for feeder position tracking
• Mantis Feeder [MF] – The mechanical and electronic unit responsible for feeding SMT tape
• Mantis Feeder Controller [MFC] – Manages communication and control, running Marlin firmware

The system supports various SMT tape sizes, including 8mm, 12mm, 16mm, 24mm, and potentially wider formats. It is designed with precision and durability in mind to ensure consistent performance.

---

Mantis Feeder PCBA: Electronic Features

The Mantis Feeder PCBA is built around an STM32 MCU running modified Photon Firmware and includes the following features:

• eFuse with reverse polarity protection and HotSwap implementation to limit the "arcing" when plugged to MFB
• 1-Wire EEPROM protection IC
• Two 2MHz constant on-time step-down switching regulators for 3.3V and 12V
• Glitch-free Power-Up and Power-Down Bus Inputs and Outputs. Low-power RS-485 transceiver which loads the RS-485 bus with only 3/20 of a standard UL, allowing more than 200 units to be connected on a single bus
• Two BDC 12V 3.6-A Peak Current Motor Drivers
• One double optical sensor for tape tension detection
• Optical quadrature sensor input from the Drive Motor encoder
• One Double/Single optical sensor for the sprocket encoder
• TC2030 SWD connection
• USART connection
• Three buttons for easy operation
• RGB Status LED

---

Mantis Feeder Controller: Communication and Control

The Mantis Feeder Controller is also based on an STM32 MCU running Marlin firmware. It includes:

• Reverse polarity protection
• Glitch-free Power-Up and Power-Down Bus Inputs and Outputs Low power RS-485 transceiver
• CAN compatibility
• Two buttons for easy operation
• Status LEDs

---

Mantis Feeder Bank: Precision and Stability

The Mantis Feeder Bank is designed and machined fully out of aluminum and stainless steel to ensure precise positioning on every plug-in and exchange of the Mantis Feeder. This is achieved through a precisely machined V-groove on the bottom plate that guides the feeder into the locking mechanism. As the feeder approaches the locking mechanism, an additional precision pin ensures accurate alignment for the main locking pin to engage securely.

The locking mechanism and its latch are spring-loaded, securing the feeder firmly against the reference plane for accurate positioning every time. Pressing the locking mechanism latch releases the feeder, automatically pushing it out for quick removal. The locking mechanism is designed to sit flush with the Mantis Feeder in the same Z-level, ensuring a smooth, unobstructed path for the Pick-and-Place machine head.

---

Mantis Feeder: Mechanical Design

The Mantis Feeder pats are designed and fully machined from aluminum, brass, stainless steel and hardened steel for high precision and durability. The aluminum machined piece, known as the Mantis Feeder Core (MFC), provides structural rigidity, while the back side of the feeder is SLS-printed for an aesthetically pleasing design. The rear housing securely holds the Mantis Feeder PCBA and Peeler Motor, along with custom Peeler brass gears and a worm gear, both supported by double bearings.

To ensure maximum precision and durability:

• All gears [total 4, two for tape advancement, and two for peeler] are custom 10-degree helix made from brass with double bearings press-fitted and secured with custom bushings for smooth operation and longevity.
• All worm gears/shafts [total 2] are custom and also made from brass, supported by two bearings, and are driven by a custom steel GM13 gearbox with a custom shaft to eliminate backlash.
• The strongest 12V motor available in this form factor drives the gearbox, providing reliable tape advancement with an additional 24ppr magnetic encoder on the back shaft of the Drive motor.
• Precision polished/vibrated laser-cut sprocket with 160 ticks per revolution, allowing 1 mm movements on its own. When combined with a quadrature encoder on the Drive motor, the achieved precision can be more than sufficient for any tape/reel size.
• An integrated leaf-spring top clamp serves a dual function:
  • It allows tape loading and unloading when the feeder is connected to the Mantis Feeder Bank. The leaf spring mechanism helps eject the top clamp, ensuring smooth loading and unloading of the SMT tape.
  • It applies pressure to the SMT tape, ensuring secure guidance for a variety of tape thicknesses.

The internal groove design accommodates different component heights:

• 8mm and 12mm SMT tape feeders feature an internal groove height of 8.5mm.
• 16mm and 24mm SMT tape feeders have a groove height of 13.5mm, allowing them to handle SMT tapes with components up to 12mm in height.

Additionally, the Mantis Feeder Bank pitch is 12.5mm, and the Mantis Feeder for 8mm tape has a width of 12mm, a dimension driven by the Drive motor.

Feeder slot allocations:
8mm Feeder - 1 slot
12mm Feeder - 2 slots
16mm Feeder - 2 slots
24mm Feeder - 3 slots

---

Summary

The Mantis SMT Feeder System is built for precision, reliability, and ease of maintenance. With robust construction, intelligent electronics, and well-designed mechanical components, it provides an efficient solution for OpenPnP users looking for a high-quality feeder system. The combination of machined aluminum, brass, stainless steel and hardened steel parts, spring-loaded locking mechanisms, and precision-machined gears ensures smooth operation and long-lasting performance.

More information soon!

Please leave a comment if this sparks your joy.

Zdenko,

[simpl...@tuta.io]

>> The Mantis SMT Feeder System is built

[Mike Menci]

Lots of people looking for feeders but nobody shows interes for this feeders .... hmm how-come ?  

[David Strum]

it looks good and well thought out.  it would be interesting to get an understanding of the order of magnitude of the cost of them.

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/openpnp/2b300811-69c0-4111-a834-fa302266a0cen%40googlegroups.com.

[Toby Dickenson]

I'm not sure how I missed this first time around. Yes, feeder options are an active interest

[j...@oxer.com.au]

I don't know why but this is the first time I noticed it, too. I can see it in the list archives but it was when Mike highlighted it that I saw it in my inbox.

My first impression is that it reads like a dream of a feeder system for OpenPnP, but the feature list (and the custom machined metal mechanicals) gives off $$$$$ vibes so perhaps that explains why it hasn't received much of a response on the list.

Another factor is that over the years many people have posted on the list about feeder projects that seemed like a dream come true, and then they fall by the wayside for one reason or another and either don't live up to the accuracy/repeatability/cost expectations or the project falls over entirely even after videos of "production ready" hardware has been shown. After seeing it happen a few times (I still mourn for the stillborn RapidFeeder) I've become jaded and don't get excited until I can see something I can order or build myself. I don't want to have my heart broken again!

I'd *love* for this to gain momentum and become a popular solution, and I hope it does. It's a very impressive project.

Jon

[Zdenko Stanec]

Greetings,

I’ll be updating this post continuously until the project is complete.
I chose not to share it earlier, even though I started over a year ago. It’s been a bumpy ride, but as always, it’s the finish line that counts the most.

The overall goal of the project was to make these feeders available as a kit, something people can easily assemble and repair themselves if necessary. The design was carefully developed with both DFA (Design for Assembly) and DFM (Design for Manufacturing) principles in mind.

A key challenge was ensuring that the price of the feeder doesn’t scale drastically with width, which is something we often see in commercial solutions. Most of the parts (about 95%) are custom-made and require MOQs, including some technical negotiations with suppliers. That’s why I don’t want users to go through the hassle of sourcing or manufacturing everything on their own, the idea is to make it simple and accessible for those who actually want to use it.

Part of the plan is also to potentially release the electronics for the feeders, as well as the modified Photon Feeder source code on my GitHub. I know how valuable open hardware and firmware can be in projects like this, so if there’s enough interest, I’d love to make those available when the time comes. This would give people the ability to build their own PCBAs and modify the source code if necessary, making it a truly flexible and customizable solution for anyone interested.

I’m trying to gauge if there’s any real interest in this kind of solution, otherwise, it will unfortunately remain dedicated solely to my own build. 

Any feedback or response is more than welcome and truly appreciated!

Sharing some of the latest tests, the machine is still WIP,

- Pick Window (Shaky Hands Edition)

- Dual Nozzle Component Pick Test

- Single Nozzle Component Pick Test

Thank you,

Zdenko,

[Jonathan Oxer]

If the question is "is anyone interested in this?" then my answer is hell yeah!

I hope you don't take my comments from my previous email in a bad way. The work you've done is very impressive, and I'd love to have feeders like this on my machines. I was just trying to explain why there may not have been a response on the list to your original post.

Please keep going, and keep us updated!

Jon

To view this discussion visit https://groups.google.com/d/msgid/openpnp/41c1d96d-cd8e-4443-9573-5332443eabfdn%40googlegroups.com.

[Zdenko Stanec]

Hey Jon,

I also noticed your initial reply in the email thread, but here it just shows "Message has been deleted." 

I've seen this happen before with Google Groups, it’s not entirely clear why, but it might be related to some spam filter or other moderation setting. The moderators might know more about the specifics.

In any case, I’ll copy your reply here since every bit of feedback is appreciated, thank you!

Zdenko,

[Dave Strum]

The feeders look awesome. 

Due to the nature of the number of feeders needed for one machine, I think price will be a large driver in the hobby community. This is obvious. 

For me (I make and sell very low volume, maybe 100 pcbs/year) anything less than about $60 id buy 4 right away to try out and buy more over time if they are as good as they appear. 60-100 id probably buy two to try and add more over time a little slower. More than $100 each would be a barrier simply because of the need for a large quantity. 

These numbers have nothing to do with the quality of the feeder itself and are entirely driven by roi for someone at my level of use. 

I hope this helps your understanding of marketability and I really hope you continue toward selling them. 

Dave

On Apr 24, 2025, at 18:04, Zdenko Stanec <zdenko...@gmail.com> wrote:



To view this discussion visit https://groups.google.com/d/msgid/openpnp/41c1d96d-cd8e-4443-9573-5332443eabfdn%40googlegroups.com.

[Jonathan Oxer]

Dave's observation about pricing (and it being based on ROI at his level of use) is very useful. Personally I'd put the numbers a little higher than that, because my use-case is slightly different. I do about 2500 PCBs / year.

This got me thinking about the cost breakdown of my current feeders. My machines are Kayo A4s converted to OpenPnP using Yamaha CL-style pneumatic feeders. Because Chinese PCBAs have moved largely from pneumatic to electric feeders in the last couple of years, the bottom has fallen out of the pneumatic feeder market and prices have dropped. The last time I bought pneumatic feeders I got a batch of 10 in August 2024 at a cost of AU$86 / unit including DHL shipping, which is about US$55 in my hands ready to go.

So for people with a machine that's already set up with the manifold, valves, etc for pneumatic feeders, you're competing with $55 / feeder.

That's going to be pretty much impossible to match for a decent electric feeder.

But the upside is that an electric feeder is mostly self-contained: apart from the controller you don't need the expensive manifold, so the initial cost of starting with just a few of them is dramatically less than the cost of setting up for pneumatic feeders where it could easily be $1k+ for the manifold, valves, pneumatic fittings, etc, not to mention a compressor. Combined with the other benefits of electric feeders, you do NOT have to beat that $55 price. For me, $100+ per electric feeder would still be a great deal if they work well. If they're narrower than the 16.5mm pitch of CL feeders, that's even better because I could load more in the machine.

Jon

To view this discussion visit https://groups.google.com/d/msgid/openpnp/812E8021-2EA6-4DC6-9E59-0850801C59E7%40gmail.com.

[Zdenko Stanec]

Hi Dave and Jon,

Thank you both very much for your valuable feedback, I really appreciate it!

The initial goal of the Mantis Feeder project was to keep the cost per feeder around €100 or less, the price I would personally be willing to pay. I won't stress enough how hard this was to achieve, given that 90–95% of the feeder consists of custom parts, made and machined according to tight specifications and tolerances. I also wanted them to be easy to operate, easy to repair, and durable. At first, my plan was to dedicate this design only to my machine, but as time went on, I saw potential to make it available for others who can benefit from them, hence my little "investigation" here to see the interest. Also, I’m happy to say that the €100 target has been achieved. Depending on how people prefer to purchase, the plan is to make options for fully assembled or partially assembled units, which could also reflect in the final price. This way, users have the opportunity to assemble the feeders themselves, which can be really helpful for understanding the mechanics for future maintenance, like lubricating the brass gears and shafts or replacing parts if needed.The feeder is designed with a focus on ease of assembly, following DFA (Design for Assembly) principles. Thanks to this approach, even a partly assembled unit can be completed quickly and without any special tools or complicated steps.

For the 3D printed parts, I chose SLS (Selective Laser Sintering) technology. The price for SLS printing has decreased significantly in recent years, making it accessible even for smaller projects. Plus, with hobby-grade SLS printers on the horizon, I didn't want to limit the design choices to what can be easily achieved with FDM printers. SLS allows for more complex, durable parts with better tolerances, making the overall product more robust.

Regarding pneumatic systems, I decided to move away from them for exactly the reasons Jon pointed out, they tend to be more expensive upfront, noisier, and require additional infrastructure like manifolds, valves, and compressors. They also struggle with flexibility, especially when dealing with smaller or larger components where multiple advances are needed to move 8, 12mm, or more. Additionally, even for 8mm tape, you often need two separate feeder types, one for 2mm advancement and another for 4mm. Not to mention the issues with components jumping, and the overall size and width of the system.

The Mantis feeder, on the other hand, offers control from 2mm to 24mm or even more in a single move across all feeder widths. It even allows for precise 1mm movement or less when combined with the quadrature encoder on the sprocket drive motor. Another addition is a visual aid/pipeline on each pick or every other pick if necessary.

There will still be an initial investment with Mantis Feeders too, mainly in the Mantis Feeder Bank (MFB), which I might bundle with the Mantis Feeder Controller (MFC). However, the plan is to keep this initial cost reasonable and definitely not anywhere near the cost of pneumatic setups. The MFB also integrates several important components like hardened bushings, laser-cut parts, springs, connectors, and special engineering plastic for durable use, but again to keep the final price accessible for hobbyists and small businesses.

To recap some details:

• Feeder pitch: 12.5mm center-to-center for 8mm tape

• Feeder widths:

  • 8mm tape → 1 slot

  • 12mm tape → 2 slots

  • 16mm tape → 2 slots

  • 24mm tape → 3 slots

• MFB scaling: Widths might be available in increments of 5 feeders (5, 10, 15, 20, etc.), depending on demand.

For example, on my machine, the Mantis Feeder Bank (MFB) on the front side is 570 mm wide, accommodating 45× 8 mm feeders with an additional 45× on the back, or 35× if I want to keep them closer to the PCB being populated. (Yes, my machine head moves freely above the feeders without any risk of crashing thanks to the feeder’s low-profile design and the 32 mm movement range of the nozzles. I also plan to make the back Mantis Feeder Bank (MFB) move with the conveyor width to save time on the machine travel, crazy! 😊)

The project is basically finished, I'm just wrapping up final details now. The 8mm and 12mm feeders will be ready first in the upcoming weeks, with 16mm and 24mm versions following soon after.

Thanks again for your encouragement and insights.

Best regards,

Zdenko,

[JW]

Hi  Zdenko,

I'm interested in what you've achieved here. When you say "will be ready first in the upcoming weeks", is that early demo videos available - or production ready, warranted units for sale with proven reliability statistics? Curious as I'm in the position of needing to invest in approx. 100 feeders in the near future, and looking at all available options right now.

[Zdenko Stanec]

Greetings JW,

These will be the first "serial production" pieces. After that, I plan to manufacture around 80 units for my own machine.

I'm still collecting feedback to decide whether I want to move forward with offering them to a wider audience. We'll see how much interest there is once I wrap up this chapter and publish a demonstration video in the coming weeks.

Zdenko,

[Toby Dickenson]

I am definitely looking forward to seeing a demonstration video.

To view this discussion visit https://groups.google.com/d/msgid/openpnp/424c3122-85d4-42c3-94d6-7fed9d27da6fn%40googlegroups.com.

[Zdenko Stanec]

Hi,

I’ve resolved the 0201 jumping issue, it took some time due to the new parts, but it’s finally done. 

Final video coming after this one.

https://www.youtube.com/watch?v=M3c5MSD6Zkw

Zdenko,

[Селенин Алексей]

Can you duplicate the video on Rutube? Otherwise, it is impossible to watch due to the blocking of SME copyrights in RU. Thank you

сб, 12 июл. 2025 г. в 21:51, Zdenko Stanec <zdenko...@gmail.com>:

To view this discussion visit https://groups.google.com/d/msgid/openpnp/2f31a5b8-c7b6-475f-9dbe-dc794526c26en%40googlegroups.com.

[Zdenko Stanec]

Hi, 

Unfortunately I can not,

Try now, it should work for you.

I will try to keep the radio music off next time,

Zdenko,

[Zdenko Stanec]

Hi, 

I made a simple tool using Python and OpenCV to capture the "initial" position of the component pocket (in this case 0201), each other pocket is measured/compared with the "initial" position (from center to center). This tool then plots the live graph and x and y "offset/error" values. This will help me understand some small details to polish the design before freezing it.

I will release this tool on my GitHub when I polish it a bit.

YT Video - So...I made a "tool"

Zdenko,

[Mike Menci]

Nice and very pro approach Zdenko - looking forward to new feeders! 

Mike 

[Clemens Koller]

Hi!

As you might notice, the x,y position is shifting around by (multiple) of 0.012mm
and no other values in between.
This might rise the question if you are hitting a limit of the resolution in your
measuring system? Maybe the image processing chain as you use it cannot process
or deliver sub-pixel information. Maybe it and can be tweaked in this regard.

Greets,

Clemens

On 03/08/2025 20.20, Zdenko Stanec wrote:
> Hi,
>
> I made a simple tool using Python and OpenCV to capture the "initial" position of the component pocket (in this case 0201), each other pocket is measured/compared with the "initial" position (from center to center). This tool then plots the live graph and x and y "offset/error" values. This will help me understand some small details to polish the design before freezing it.
>
> I will release this tool on my GitHub when I polish it a bit.
>

> YT Video - So...I made a "tool" <https://www.youtube.com/watch?v=A5YvtPbbksk>

>
> Zdenko,
>
> On Saturday, 12 July 2025 at 21:45:55 UTC+2 Zdenko Stanec wrote:
>
> Hi,
>
> Unfortunately I can not,
>
> Try now, it should work for you.
>
> I will try to keep the radio music off next time,
>
> Zdenko,
>
> On Saturday, 12 July 2025 at 20:58:10 UTC+2 sveto...@gmail.com wrote:
>
> Can you duplicate the video on Rutube? Otherwise, it is impossible to watch due to the blocking of SME copyrights in RU. Thank you
>
> сб, 12 июл. 2025 г. в 21:51, Zdenko Stanec <zdenko...@gmail.com>:
>
> Hi,
>
> I’ve resolved the 0201 jumping issue, it took some time due to the new parts, but it’s finally done.
>
> Final video coming after this one.
>

> https://www.youtube.com/watch?v=M3c5MSD6Zkw <https://www.youtube.com/watch?v=M3c5MSD6Zkw>

>
> Zdenko,
>
> On Monday, 5 May 2025 at 19:02:31 UTC+2 to...@tarind.com wrote:
>
> I am definitely looking forward to seeing a demonstration video.
>
> On Mon, 5 May 2025 at 08:43, Zdenko Stanec <zdenko...@gmail.com> wrote:
>
> Greetings JW,
>
> These will be the first "serial production" pieces. After that, I plan to manufacture around 80 units for my own machine.
>
> I'm still collecting feedback to decide whether I want to move forward with offering them to a wider audience. We'll see how much interest there is once I wrap up this chapter and publish a demonstration video in the coming weeks.
>
> Zdenko,
> On Monday, 5 May 2025 at 04:01:48 UTC+2 JW wrote:
>
> Hi Zdenko,
>
> I'm interested in what you've achieved here. When you say "will be ready first in the upcoming weeks", is that early demo videos available - or production ready, warranted units for sale with proven reliability statistics? Curious as I'm in the position of needing to invest in approx. 100 feeders in the near future, and looking at all available options right now.
>
> On Saturday, April 26, 2025 at 10:33:35 AM UTC+1 zdenko...@gmail.com wrote:
>
>
> Hi Dave and Jon,
>
> Thank you both very much for your valuable feedback, I really appreciate it!
>
> The initial goal of the Mantis Feeder project was to keep the cost per feeder around €100 or less, the price I would personally be willing to pay. I won't stress enough how hard this was to achieve, given that 90–95% of the feeder consists of custom parts, made and machined according to tight specifications and tolerances. I also wanted them to be easy to operate, easy to repair, and durable. At first, my plan was to dedicate this design only to my machine, but as time went on, I saw potential to make it available for others who can benefit from them, hence my little "investigation" here to see the interest. Also, I’m happy to say that the €100 target has been achieved. Depending on how people prefer to purchase, the plan is to make options for fully assembled or partially assembled units, which could also reflect in the final price. This way, users have the opportunity to assemble the feeders themselves, which can be really helpful for
> understanding the mechanics for future maintenance, like lubricating the brass gears and shafts or replacing parts if needed.The feeder is designed with a focus on ease of assembly, following DFA (Design for Assembly) principles. Thanks to this approach, even a partly assembled unit can be completed quickly and without any special tools or complicated steps.
>
> For the 3D printed parts, I chose SLS (Selective Laser Sintering) technology. The price for SLS printing has decreased significantly in recent years, making it accessible even for smaller projects. Plus, with hobby-grade SLS printers on the horizon, I didn't want to limit the design choices to what can be easily achieved with FDM printers. SLS allows for more complex, durable parts with better tolerances, making the overall product more robust.
>
> Regarding pneumatic systems, I decided to move away from them for exactly the reasons Jon pointed out, they tend to be more expensive upfront, noisier, and require additional infrastructure like manifolds, valves, and compressors. They also struggle with flexibility, especially when dealing with smaller or larger components where multiple advances are needed to move 8, 12mm, or more. Additionally, even for 8mm tape, you often need two separate feeder types, one for 2mm advancement and another for 4mm. Not to mention the issues with components jumping, and the overall size and width of the system.
>
> The Mantis feeder, on the other hand, offers control from 2mm to 24mm or even more in a single move across all feeder widths. It even allows for precise 1mm movement or less when combined with the quadrature encoder on the sprocket drive motor. Another addition is a visual aid/pipeline on each pick or every other pick if necessary.
>
> There will still be an initial investment with Mantis Feeders too, mainly in the Mantis Feeder Bank (MFB), which I might bundle with the Mantis Feeder Controller (MFC). However, the plan is to keep this initial cost reasonable and definitely not anywhere near the cost of pneumatic setups. The MFB also integrates several important components like hardened bushings, laser-cut parts, springs, connectors, and special engineering plastic for durable use, but again to keep the final price accessible for hobbyists and small businesses.
>
> To recap some details:
>

> *
>
> *Feeder pitch*: 12.5mm center-to-center for 8mm tape
>
> *
>
> *Feeder widths*:
>
> o
>
> 8mm tape → 1 slot
>
> o
>
> 12mm tape → 2 slots
>
> o
>
> 16mm tape → 2 slots
>
> o
>
> 24mm tape → 3 slots
>
> *
>
> *MFB scaling*: Widths might be available in increments of 5 feeders (5, 10, 15, 20, etc.), depending on demand.
>
> For example, on my machine, the Mantis Feeder Bank (MFB) on the front side is 570 mm wide, accommodating 45× 8 mm feeders with an additional 45× on the back, or 35× if I want to keep them closer to the PCB being populated. /(Yes, my machine head moves freely above the feeders without any risk of crashing thanks to the feeder’s low-profile design and the 32 mm movement range of the nozzles. I also plan to make the back Mantis Feeder Bank (MFB) move with the conveyor width to save time on the machine travel, crazy! 😊)/

>
> The project is basically finished, I'm just wrapping up final details now. The 8mm and 12mm feeders will be ready first in the upcoming weeks, with 16mm and 24mm versions following soon after.
>
> Thanks again for your encouragement and insights.
>

>> - Pick Window (Shaky Hands Edition) <https://youtu.be/B-0Q9Bfb2sQ>
>>
>> - Dual Nozzle Component Pick Test <https://youtu.be/sVIBBZIJoCI>
>>
>> - Single Nozzle Component Pick Test <https://youtu.be/G7TMQRGlqes>

>>
>> Thank you,
>>
>> Zdenko,
>>
>> On Wednesday, 23 April 2025 at 00:57:48 UTC+2 j...@oxer.com.au wrote:
>>
>> I don't know why but this is the first time I noticed it, too. I can see it in the list archives but it was when Mike highlighted it that I saw it in my inbox.
>>
>> My first impression is that it reads like a dream of a feeder system for OpenPnP, but the feature list (and the custom machined metal mechanicals) gives off $$$$$ vibes so perhaps that explains why it hasn't received much of a response on the list.
>>
>> Another factor is that over the years many people have posted on the list about feeder projects that seemed like a dream come true, and then they fall by the wayside for one reason or another and either don't live up to the accuracy/repeatability/cost expectations or the project falls over entirely even after videos of "production ready" hardware has been shown. After seeing it happen a few times (I still mourn for the stillborn RapidFeeder) I've become jaded and don't get excited until I can see something I can order or build myself. I don't want to have my heart broken again!
>>
>> I'd *love* for this to gain momentum and become a popular solution, and I hope it does. It's a very impressive project.
>>
>> Jon
>>
>> On Wednesday, 23 April 2025 at 5:50:37 am UTC+10 to...@tarind.com wrote:
>>
>> I'm not sure how I missed this first time around. Yes, feeder options are an active interest
>>
>> On Tue, 22 Apr 2025, 20:15 Mike Menci, <mike....@gmail.com> wrote:
>>
>> Lots of people looking for feeders but nobody shows interes for this feeders .... hmm how-come ?
>>
>> On Friday, 11 April 2025 at 23:21:09 UTC+2 simpl...@tuta.io wrote:
>>

>> /*>> The Mantis SMT Feeder System is built*/

>>
>>
>>
>> --
>> You received this message because you are subscribed to the Google Groups "OpenPnP" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

>> To view this discussion visit https://groups.google.com/d/msgid/openpnp/2b300811-69c0-4111-a834-fa302266a0cen%40googlegroups.com <https://groups.google.com/d/msgid/openpnp/2b300811-69c0-4111-a834-fa302266a0cen%40googlegroups.com?utm_medium=email&utm_source=footer>.

>>
>> --
>> You received this message because you are subscribed to the Google Groups "OpenPnP" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

>> To view this discussion visit https://groups.google.com/d/msgid/openpnp/41c1d96d-cd8e-4443-9573-5332443eabfdn%40googlegroups.com <https://groups.google.com/d/msgid/openpnp/41c1d96d-cd8e-4443-9573-5332443eabfdn%40googlegroups.com?utm_medium=email&utm_source=footer>.

>
> --
> You received this message because you are subscribed to the Google Groups "OpenPnP" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
>

> To view this discussion visit https://groups.google.com/d/msgid/openpnp/812E8021-2EA6-4DC6-9E59-0850801C59E7%40gmail.com <https://groups.google.com/d/msgid/openpnp/812E8021-2EA6-4DC6-9E59-0850801C59E7%40gmail.com?utm_medium=email&utm_source=footer>.

>
> --
> You received this message because you are subscribed to the Google Groups "OpenPnP" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
>

> To view this discussion visit https://groups.google.com/d/msgid/openpnp/424c3122-85d4-42c3-94d6-7fed9d27da6fn%40googlegroups.com <https://groups.google.com/d/msgid/openpnp/424c3122-85d4-42c3-94d6-7fed9d27da6fn%40googlegroups.com?utm_medium=email&utm_source=footer>.

>
> --
> You received this message because you are subscribed to the Google Groups "OpenPnP" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
>

> To view this discussion visit https://groups.google.com/d/msgid/openpnp/2f31a5b8-c7b6-475f-9dbe-dc794526c26en%40googlegroups.com <https://groups.google.com/d/msgid/openpnp/2f31a5b8-c7b6-475f-9dbe-dc794526c26en%40googlegroups.com?utm_medium=email&utm_source=footer>.

>
> --
> You received this message because you are subscribed to the Google Groups "OpenPnP" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com <mailto:openpnp+u...@googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/openpnp/bfc82cc8-af5c-4bcb-9644-07f52553dfb6n%40googlegroups.com <https://groups.google.com/d/msgid/openpnp/bfc82cc8-af5c-4bcb-9644-07f52553dfb6n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Селенин Алексей]

Не знаю, по моему удалилось сообщение почему-то. Привет Zdenko. Твоя новая машина просто монстр! Поздравляю! Машину первой версии ты разобрал наверное? Мне нравится концепция. Я хочу повторить ее. Прошу прощения, не смог бы ты поделится чертежами алюминиевых деталей своей первой машины которая на валах SFC? У нас в маленьком городе невозможно заказать детали CNC. Буду искать в Китае возможно. Спасибо. С ув. Алексей.

вс, 3 авг. 2025 г. в 21:20, Zdenko Stanec <zdenko...@gmail.com>:

To view this discussion visit https://groups.google.com/d/msgid/openpnp/bfc82cc8-af5c-4bcb-9644-07f52553dfb6n%40googlegroups.com.

[Zdenko Stanec]

Hi,

@Clemens – The position measurement is directly proportional to unit/pixel. The test was performed on a 0201 package, which is quite small, the measuring accuracy is pretty good for this setup, but there is always room for improvement.

@Svetozar – Thank you for the kind words. The first machine has found a new owner. Send me a direct email, and I’ll see what I can do.

Zdenko,

[Селенин Алексей]

Thank you Zdenko. Very glad to hear this. My mail sveto...@gmail.com. With regards Alexey

ср, 27 авг. 2025 г. в 09:30, Zdenko Stanec <zdenko...@gmail.com>:

To view this discussion visit https://groups.google.com/d/msgid/openpnp/03af4af0-9f47-430a-85d1-20de6d45fab1n%40googlegroups.com.

[Zdenko Stanec]

Hi,

I wanted to share the final stage of development. It has been quite a ride, but I am now closing all remaining open points. The parts will be sent for anodization in about two weeks, and after that I will have the finished product in my hands.

Some additional features implemented:

Feeder protocol and communication performance :

The feeder runs on the standard PhotonFeeder protocol, but the firmware on the Mantis Feeder has been completely rewritten to support the new mechanical and HW design.

Normally, the PhotonFeeder protocol runs over Marlin, which introduces communication delays. To solve this, I developed the Mantis Bridge, a small PCBA with custom firmware that enables extremely fast communication between OpenPnP and the Mantis Feeder.

How Fast is it?

From the moment OpenPnP sends the Feed command, the Mantis Feeder receives it, processes the movement, feeds, and replies back in ~80 ms for 2 mm pitch.

This translates to roughly 12 components/second.

Every additional 2 mm of pitch adds only ~50 ms, meaning for 4 mm pitch the full cycle is ~130 ms, and so on.

RS485 Firmware Update Support :

The Mantis Feeder supports firmware updates over RS485.

A custom bootloader was developed to allow FW updates via RS485 as well as One-Wire EEPROM programming which are located on Mantis Feeder Bank (MFB).

STM32 RS485 Firmware Manager tool was developed which allows this communication. It detects all feeders, loads new FW, and programs them one by one supporting up to 250 feeders on a single bus.

Injection-molded sprockets :

Since the project progressed so far, the final decision was to manufacture injection-molded sprockets to ensure long-term stability and reliability eliminating any production issues such as in laser cut sprockets, etc...(still WIP, in communication with supplier)

Modular reel carrier design :

The reel carrier is designed to be interchangeable and adaptable for different machine types.

You can also use the feeder without the reel holder if you prefer placing the feeder closer to the PCB to reduce machine travel time, and mount the reels elsewhere on the machine

 

Additional photos:

 

Thank you for following this build , I hope you like it.

Have a lovely weekend,

Zdenko,

P.S.

I have more ideas for this ecosystem in the future, but first I need to close this chapter.

If you'd like to help or contribute in any form, your support would be greatly appreciated.

[Mike Menci]

Well done Zdenko ! 🪫 

Take some holidays now and recharge your batteries 🔋 

Might come to visit you one day - you have many your toys are sparkling me! 

Cheers! 

Dne 1. nov. 2025 ob 11:34 je oseba Zdenko Stanec <zdenko...@gmail.com> zapisala:



Hi,

I wanted to share the final stage of development. It has been quite a ride, but I am now closing all remaining open points. The parts will be sent for anodization in about two weeks, and after that I will have the finished product in my hands.

Some additional features implemented:

Feeder protocol and communication performance :

The feeder runs on the standard PhotonFeeder protocol, but the firmware on the Mantis Feeder has been completely rewritten to support the new mechanical and HW design.

Normally, the PhotonFeeder protocol runs over Marlin, which introduces communication delays. To solve this, I developed the Mantis Bridge, a small PCBA with custom firmware that enables extremely fast communication between OpenPnP and the Mantis Feeder.

How Fast is it?

From the moment OpenPnP sends the Feed command, the Mantis Feeder receives it, processes the movement, feeds, and replies back in ~80 ms for 2 mm pitch.

This translates to roughly 12 components/second.

Every additional 2 mm of pitch adds only ~50 ms, meaning for 4 mm pitch the full cycle is ~130 ms, and so on.

<log.jpg>

RS485 Firmware Update Support :

The Mantis Feeder supports firmware updates over RS485.

A custom bootloader was developed to allow FW updates via RS485 as well as One-Wire EEPROM programming which are located on Mantis Feeder Bank (MFB).

STM32 RS485 Firmware Manager tool was developed which allows this communication. It detects all feeders, loads new FW, and programs them one by one supporting up to 250 feeders on a single bus.

<Tool.jpg>

Injection-molded sprockets :

Since the project progressed so far, the final decision was to manufacture injection-molded sprockets to ensure long-term stability and reliability eliminating any production issues such as in laser cut sprockets, etc...(still WIP, in communication with supplier)

Modular reel carrier design :

The reel carrier is designed to be interchangeable and adaptable for different machine types.

You can also use the feeder without the reel holder if you prefer placing the feeder closer to the PCB to reduce machine travel time, and mount the reels elsewhere on the machine

<WhatsApp Image 2025-11-01 at 10.17.35_a32d6e59_1440x1080.jpg>

<WhatsApp Image 2025-11-01 at 10.16.58_66f13ce6_1440x1080.jpg>

 

Additional photos:

<WhatsApp Image 2025-11-01 at 10.22.52_51fb3e63_1440x1080.jpg>

<Screenshot 2025-11-01 102836.jpg>

To view this discussion visit https://groups.google.com/d/msgid/openpnp/edb23804-3f94-400d-96f5-83fe135dea62n%40googlegroups.com.

<log.jpg>

<WhatsApp Image 2025-11-01 at 10.16.58_66f13ce6_1440x1080.jpg>

<Tool.jpg>

<WhatsApp Image 2025-11-01 at 10.22.52_51fb3e63_1440x1080.jpg>

<Screenshot 2025-11-01 102836.jpg>