Rhino Smart: Extruder Jam Detector for 3D printers
Never worry about dry running your 3D printer again. RhinoSmart will pause the print if printer runs out of filament or there is a jam
The Story behind
I started MONIRAD Robotics over 2 years ago. With support of over 100 Kickstarter backers we started manufacturing of the Rhino Industrial grade 3D printer around March 2015. I have been developing RhinoPrinter ever since, improving the quality of prints and increasing the reliability. One of the main concerns specialty during a day long print job is running out of filament before the print ends. In case of a small part it is easy to estimate whether the remaining filament in the spool is enough to finish the print but if you have a large print you need to check the remaining filament every now and then to make sure the printer is not running dry. This could be a problem especially if you want to run the printer overnight.
A very simple solution is to use an on-off or an optical switch to detect the presence of filament and pause the print in case the filament runs out. However based on 2 years of experience working with 3D printers another likely case of ending up with a half-finished printed part is when the extruder is jammed.
Jamming could occur due to one of the following.
- The filament gets tied in the spool holder hence the extruder can't pull it thought.
- Both ABS and PLA filaments absorb the moisture in the air. Excessive moisture in the filament could cause a jam in the nozzle specially when using a 0.3mm or smaller nozzle size during the print.
- The hotend gets clogged due to lack of sufficient cooling on the hotend heatsink.
- The nozzle is too close to the bed during printing the first layer. Or filament diameter is more than its nominal value causing formation of a convex solid layer. As a result there isn’t enough gap between nozzle tip and printed surface during printing the next layer.
- Extruder can’t keep up with print speed.
Due to the above reasons I was looking for a module on my printers that could detect both out of filament and jamming, unfortunately I noticed there isn't much development in this area. There are products that detect out of filament but fail to pause the print if a jam occurs. Some companies have developed this module but it is compatible only to their own 3D printer brand.
So I went through couple of mechanical design iterations to get an inexpensive and reliable system to measure the amount of extruded filament. Then I completely rewrote the algorithm for detecting a jam for a couple of times until I got a very reliable system to ensure it is not too sensitive and is capable of detecting the jam for small or large prints.
Having feedback from the extruders is a necessity. Hence the goal in designing Rhino Smart was to make it available at an affordable price. The rotary encoder is a standard Arduino switch reinforced with a ball bearing for increased durability and life cycle. The body consists of five 3D printed parts made out of ABS and printed in a printer with heated chamber which would considerably increases strength of printed parts. We will also include 3 sets of wires that allows you to adjust the wire length between 20cm to 80cm.
The Electronic Board is a Arduino Rotary Encoder with built in switch. The pins are: Clock, Direction, Switch, +5V and Ground. If over 110 supporter request for Rhino Smart Developer Edition we will design a PCB board with dedicated microcontroller that can communicate with host via UART, SPI and I2C protocol.
At the moment Rhino Smart is available for RhinoPrinter board which runs on Teacup firmware. It is supported by repetier firmware and with your support I would be able to expand the firmware compatibility to Marlin, Smoothie and other popular platforms. I will publish the pseudo code as well as sample code written in plain C for developers, making it possible for everyone to contribute in developing the firmware. Upon successful campaign I will also send sample units to official firmware developers asking them to embed the Rhino Smart driver in the official firmware release.
To ensure there will be a support for a specific firmware I dedicated a separate reward for each firmware. so once you pledge you can see how many supporters are using a specific firmware. If you can't see your firmware in the reward section please send me a message and I will add it to the list. If the number of supporters reach a minimum of 100 backers then we will make sure that the hardware is supported by the firmware. upon completion of the campaign the rewards with less than 110 backers can ask for a refund we can't promise compatibility of the module with that specific firmware.
Production cost and Delivery time
I made 10 units of Rhino Smart and used the total assembly time as a basis for estimating the labor cost. Based on the experience from the previous KickStarter Campaign I include all the additional expenses that are required for on time completion of the project. So we estimate the total cost of each unit including the cost for firmware development to be around 30USD. on the top of this cost Kick Starter takes a 5% commission and there is a 3% bank transfer fee. We will also loose a small some during USD to CAD conversion and back from CAD to USD for paying for the raw materials. On the other hand we save a bit on the labor cost considering USD to CAD conversion rates.
The cost breakdown is presented the below table.
We anticipate each unit to take around 3 hours of print time and with dedicating six 3D printers for printing the parts we will be able to produce enough parts for 16 units per day if we work 8 hours/day. Since the manual assembly and testing of each unit will take around half an hour, we should be able to manufacture minimum of 16 units/day hence we anticipate production of 400 units/month based on 40 hours/week of work. The lead time for ordering the parts is less than a month and we anticipate to be able to develop the driver for each firmware in less than two month including the beta testing of the unit. Upon completion of the campaign it takes around 3 weeks to receive the funds from Kickstarter hence we will be sending out 400units/month starting 15th of March 2017.
Risks and challenges
We have gained valuable experiences from our first Kickstarter campaign. That is why before the campaign I printed 10 additional prototypes to get a better estimate of risks and challenges.
The problems that could delay the productions are
A) During assembly of the first 10 units I noticed one out of 10 encoders were defected. A defected encoder could slow down the production rate by wasting the time spent on the assembly of the unit and additional wait time for the arrival of raw material. Our goal is to order 1.2 times more the raw material needed to fulfill all the orders to avoid any delay in production. If the ratio of defected parts exceed 20% that would cause a maximum of a month delay on fulfilling the orders.
B) running out of raw materials ( Arduino switches ball bearing and 3D printer filaments) hence we assume one month lead time to order all the raw material needed to manufacture the entire units upon successful campaign
C) Failing to recruit skilled labor capable of operating and debugging 3D printers. Especially as it would be based on a short term contract. Solution: I have already discussed this potential job with the previous employees that are familiar with assembly of RhinoPrinter.
D) We have a contract with UPS to get a discounted rates with UPS, so there would be no suppressed shipping cost for the customers in North America. On the other hand we need to renew our contract with DHL for international shipments. That could further delay overseas shipments.
E) Development of the custom firmware for a less popular motherboard could be an issue. i.e. you may already receive the hardware but you would have to wait for the firmware to be developed for you specific motherboard running on a specific firmware.
F) In case we reach our stretch goal and decide to offer the mold injected parts the lead time to prepare the mold is between 46 to 60 days. In addition to that I would need to locate a warehouse for storage and assembly of the units. This would increase the delivery of the first batches by 2 to 3 months. However we will be able to increase the manufacturing rate by orders of 10 so this extra step will not affect delivery time for the supports that are scheduled for delivery on or after mid-June 2017.
- (43 days)