We ran into an issue in our certification screening last week – we’re putting this update out to keep you in the loop with everything. Certification is a tricky game, we’ll get into some of the details below.
Now on to the details….
We mentioned in our January post that we were headed in for a testing session with a company who we’d be working with to certify The Palette. Part of this was to use a super sensitive antenna to measure any electromagnetic signals coming off of The Palette and its power supply – you can see a picture of the test rig below.
In this update we’ll run you through the certification process, the decisions we have made, what the certification obstacles with The Palette are, and what our plan of action is.
You may (or may not) know that many products in our field (consumer 3D printing) are shipped uncertified, so why certify at all?
We briefly considered this option, but decided that because we’re working with a new product that will one day be in thousands of people’s homes – we’re going to bite the bullet and get everything certified to the necessary standards. We want to make sure we dot our i's and cross our t's before we ship you your Palette. We don’t cut corners when it comes to quality, safety, or transparency and we are not going to in this area either.
Something to understand is that even though The Palette did not pass its initial certification run does not mean that it is in any way unsafe. Government bodies, such as the FCC, have requirements that a company’s product must meet to ensure it does not interfere with any signals, or other products (i.e. radios, remote controls etc). These emitted frequencies are where the majority of our certification issues are (more on this below).
You might be asking yourself why we didn’t choose to go out months ago and certify The Palette. The main purpose of a product being certified is to ensure that the final version meets certain specifications. There is a very important word in that last sentence – final. Final product means that once we have it certified, it cannot go through any changes, or we must go through the entire process again.
This finality is the reason we were not able to pursue certification months ago.
We believe in the importance of preparation, planning, and mitigating risks. We’re going to lay out some of the steps we’ve taken over the preceding few months to try and achieve certification on the first go.
As with many things, the output (product) is only as good as the inputs (components) that go into it. With this in mind, we sourced certified components where available. The most important of these components when it comes to certification is The Palette’s power supply – a UL certified part that comes from a reputable supplier.
This power supply failed the certification test. We’re working with a Class 2 supply, which means we need to get The Palette and the power supply certified together. We had The Palette tested, and came to discover a few additional problems, but once we saw the UL certified power supply failed we knew it wasn’t going to be a fun day.
Below is a graph from the certification pre-screen, we’ll talk you through what this means in the following paragraphs.
There are three lines on this graph – one is blue, one is green and one is red.
The red line is the reading measured from The Palette at different frequencies. The blue line is the level you must be below for industrial certification. The green line is the level you must be below for consumer certification.
You can see the challenge lies in the low frequency range (30-70Mhz) where The Palette was above both the green and blue lines. The experts we worked with at the certification facility call this “low frequency noise”.
What does this mean in English? What components are causing what problems?
The good news is, we know where the problems are coming from. This low frequency noise is caused largely by the power supply we mentioned above. The spikes in the middle are caused by the Palette’s Arduino/RAMPS setup.
Unlike mechanical systems where you can usually pinpoint a problem using your senses, these emissions problems are much more difficult to determine the source(s) of the problem. The experts we worked with told us that EMC is like black magic and that a lot of things cannot be solved analytically. They explained that sometimes things as simple as shortening a wire can have a huge affect on the emission performance. It took extensive systematic testing – variable by variable – to identify areas that are contributing to emissions and areas that are not.
So we know there are some issues, we have an idea as to where those issues are coming from, what are we doing about it?
We’ve brought in an expert EMC consultant to help us identify and implement solutions. We believe this process will move much quicker by leveraging someone who has solved similar issues before. We got a few references from our network, and completed our first design session earlier this week.
To test the electromagnetic emissions of a product, you need to do it in a special chamber like the one shown above where we tested The Palette. There are only a few in Ontario and renting them is very expensive. This delays the development cycle, as we need to book sessions in the chamber every time we need to implement and validate a change. This process is systematic, thorough and slow by nature. For these reasons we chose to invest in an EMC consultant to help expedite this critical process as much as possible.
Unfortunately, this has put a delay on our manufacturing plans. We’re going to continue ordering parts (everything but the power supply), but we cannot move forward into any assembly, or shipping until this certification issue has been solved. We have sourced several alternative power supplies that we will verify for use with The Palette during our next emission testing session. Once we select a suitable supply we will arrange for an expedited bulk order.
We thought about assembling the first hundred Palettes, and waiting until the product passed certification but decided the risk, and associated cost, was too high if the hardware inside The Palette needed to be changed.
We’ve started getting more and more parts in, and we’re building more test units. The production team plans to use the newfound time to continue to refine our assembly processes and to implement further quality control checks. The development team will similarly get more hours of testing in to help hammer out bugs and build out features.
Overall, the team is excited to push through this last unknown so we can get Palettes out to all corners of the world.