Why we (usually) don’t make expanders
Frequently, customers will ask for additional functionality added to a module with the use of an expander. If you’re unfamiliar, an expander is a secondary module that connects to a main module behind the panel and adds additional connections and/or functionality. In the early days of Eurorack, and, in fact, in the early days of Noise Engineering, this was a very common practice (see Mimetic Sequent and Mimetic Multium, for example); in recent years it’s fallen slightly out of favor, but you’ll still see expanders pop up from time to time. They’re a fun concept since they can add even more flexibility and customization to an already fluid format.
So, why don’t we make them much these days? In this post, we’ll share a bit behind the scene about why they present challenges, why we generally don’t make them any more,
The most common reason: We can’t
Most frequently, when an expander is requested to add new functionality to an existing module, we simply can’t. Expanders require dedicated hardware designed into the module from the start: expanders require dedicated headers that interact with the rest of the circuitry on a module, so if we didn’t design a header in when a module was produced, there’s no way to add additional hardware functionality.
Another big reason: Market size and production
Making an expander for an existing product is generally impossible, but designing one beforehand is almost as challenging for a completely different set of reasons. It all has to do with numbers.
Eurorack is a relatively niche product with a small customer base (compared to fixed-architecture synthesizers, for example). If we design a Eurorack module that we like, though, we can expect to sell a certain number of said module: everyone with a Eurorack case has the option of buying our new product.
With an expander, though, that arithmetic changes. Now, we have an expander that can only be purchased by customers who have already purchased one specific other product.
This is further complicated by the logistics of production. We can’t just order a single module from a manufacturer: there are minimum order quantities to keep costs down, and production takes time. In general, we can keep costs down by ordering intelligently: we do some projections, evaluations, and planning (aka we make it up as we go and hope for the best -kris) to make sure that a new product is viable and ordered in the correct quantity. At this point, we’ve gotten reasonably good at the calculations to make sure we’re not way off on our production quantities.
With expanders, that math goes out the window. Maybe every customer that buys X module will buy the expander… or maybe half will, or maybe a quarter will. Since there are minimum order quantities for a production run, we can’t always build the most conservative number that we think will sell. And if we run out of stock of X module but still have its expander in stock, they generally sit forever. A real life example: at time of writing we still have a couple of Mimetic Sequent expanders, and that module was discontinued in 2020.
Design
As we’ve matured, we’ve gotten more comfortable with making concrete design decisions. In the early days, expanders were an easy way to add in features that might be desirable, or might not. This taught us a lot, but these days, we have a much better idea of how we want our products to work in a patch. We often design subtractively: we put together every feature that we might want on a module, then chop it down to a core design from there. Size, cost, and desired functionality all play into that decision, but we can usually come to a consensus on how we want a final module to work before we finalize the design.
Where it CAN make sense
You’ll notice that we do have one module that does have an expander: Xer Mixa and its friend Expando Expandi. There were some very careful calculations done to make sure that was the right choice, and it has played out how we hoped – we actually have had a hard time keeping both modules in stock.
Xer Mixa has extensive possibilities for what it can do internally. This meant that we could support a huge number of CV inputs if we wanted to. The core module is small enough to fit in a performance case, and in a system like that, HP comes at a premium. It’s also more than powerful enough to be the mixer for a monster rack, and in a system like that, having 8 or even 16 freely routable CV inputs is a huge boon.
We also knew that, from the start, Xer Mixa was going to be expensive… and Expando Expandi was not. The extensive analog section on Xer Mixa really adds up, and the 8 jacks and minimal electronics on Expando Expandi makes it quite simple by comparison. We did the math and figured out that we could offer an attractive discount by selling the pair of them together, which made it a much easier expander to sell than previous designs. And, since multiple Expando Expandi can be connected to one Xer Mixa, there’s already a higher market potential than there is for something like our beloved Variatic Sequent.
What about USB expanders for Alia, Versio, and Legio?
We’ve also gotten a lot of requests for a USB breakout for our platform modules over the years. Initially, our firmware update process required you to press the two buttons on the back of the module to program it, so a USB expander didn’t make much sense. Recently, though, we updated that process so that no button pressing is required to change between official firmwares, and we immediately decided to do the expander, going so far as buying some sample cables and milling out a panel for it. When we started exploring how to make the connection between the front and the back of the panel, things got complicated.
First, Legio has a different USB placement than Versio and Alia, due to its smaller size. It points up toward the top of the module, and it’s near the top of the board. In most cases, you would need an angled adapter to fit into it. While an angled connector makes the most sense for Legio, that either makes the module incredibly deep if it angles back (adding a couple of inches behind the panel), or makes the module wider than 6 HP if it angles to the side. A straight connection would require clearance between the cable and the rail, which is very tough to specify and won’t work in a majority of cases.
We’re also working to move all of our modules to USB-C: it’s become quite ubiquitous, and we prefer it overall to micro USB.
This means that in order to create USB breakout modules, we would need four versions: micro USB for existing modules, USB C for new stock, angled cables for Legio, and straight cables for Versio and Alia.
For these reasons, we’ve decided not to offer an official version. In our systems, we generally just leave a gap and run a USB cable out from behind the modules – this takes up less space anyway – but if you want to go the DIY route, some other companies have made breakouts that may work with some careful measurements and cable adapters.
What else are you curious about?
This was one of the most common questions we’ve gotten over the years, so we figured it was a good question to address in a longer-form post. Behind the scenes, Noise Engineering is just four people, and being a small business in an industry like Eurorack is a unique experience. If you have more behind-the-scenes questions, drop us a line – we’d love to make this a series.
