Metal or not metal? That is the question!

This Hamletesque question has haunted our team in connection with Omnia for a few months. Turris Omnia was introduced as a home router in a nicely shaped plastic case and for a long time we did not even think of other options. 5 GHz Wi-Fi connection was intended to be provided by three outside antennas and the “older” Wi-Fi at the 2.4 GHz frequency was supposed to be broadcast, sort of obligatorily, with two internal antennas, more or less for backwards compatibility with older devices.

This whole project aroused great interest. At the beginning, we couldn’t even dare to hope for our current 2,500+ supporters who contributed the incredible 530 thousand USD. Thanks to them, a very interesting discussion started on Indiegogo, where many people asked us to make a metal version of the case. On the other hand, later their opponents appeared, who felt that a plastic case was better quality.

Unfortunately, production of both plastic and metal cases doesn’t make too much sense, especially when it comes to thousands of pieces. This is because the production of both is, unfortunately, conflicting. The plastic case has a high mold development cost and a relatively cheap subsequent production, while the development of the metal case is not as costly, but the production costs are higher. So, if we went for both versions, it would bring both their negative impacts: we’d pay the high costs of mold development and the high production costs of the metal version. That is why it is not possible to offer both.

Therefore, we debated over the qualities of the two versions and considered which of them would fit best for our Omnia:

1) Appearance
I should warn you that I’m a technical enthusiast, geek, nerd who certainly does not follow the latest trends and never even asked himself what color he wants his router, server, etc. to be. However, you may be surprised that from my point of view the metal (more precisely, the anodized aluminum) version looks much better than the plastic one. While it is true that if I compare the published photos, I rather like the plastic, but in reality plastic looks worse than in the picture and, on the contrary, metal looks much better in real life than in the picture that we were able to provide in this phase of the project.

2) Cooling
From the beginning we promised that Turris Omnia will be one of the fastest in its class, and we are still pursuing this target vigorously. Unfortunately, powerful chipsets often generate more heat, which in turn increases the cooling requirements. In the metal case, we could benefit from the distribution of heat from the board to the case body, which then acts as a large passive heat sink. Plastic, unfortunately, is a heat insulator, which makes cooling more complicated. When testing the prototypes of the plastic case in extreme conditions of long-term maximum load, we even came across the need of software-based restriction of the processor performance, in order to avoid overheating. So there was no margin for heat capacity that we desired, especially considering the possibility of adding more cards that would also act as sources of heat. (Of course, we have rejected the possibility of active cooling already in the beginning.) Therefore, metal scored in this one, too.

3) Antennas
Antennas might look like the weakest spot of the metal case, but only at first glance. From the physical point of view, it is simply impossible for a small internal antenna suspended just a few centimeters above the top of the router’s board full of copper and perhaps hidden under a large heat sink to give the same performance as a large external antenna. How can you then solve the problem of 3 + 2 antennas for Wi-Fi? There are two solutions: either to lay all five antennas out, as we did in the case of our prototype, or to use dual antennas for both bands that will merge the signals for both frequencies. We plan to use the same signal combiners that are commonly used in Wi-Fi cards supporting both frequencies.

4) Extensibility
It might seem that as far as extensibility is concerned, the metal solution would lose to the plastic one. But on the other hand, the rear panel in the metal solution will contain a sufficient amount of openings for LTE modems, GPS antennas, etc., so no handicap would arise. In addition, production of metal cases is far more flexible, so we can offer special cases with extra openings for various applications.

5) Time to market
A big advantage of a metal construction is that it doesn’t require the development of mold. That is a fairly lengthy process and with regard to the ambitious goals we set for the Turris Omnia project, this process could be too lengthy. If we once again reverted to the plastic version in the current phase of the project, we would find it difficult to comply with the promised production deadlines. And we assume that you are looking forward to your new piece of hardware as much as we are, and each day of delay would be as long for you as it’d be for us.

6) The overall appearance
Our team is in touch with prominent managers of large producers of commodity and security hardware, so we understandably didn’t miss the opportunity to consult the matter with them. And they convinced us that a product like Omnia definitely deserves a better and more robust sheet metal chassis. If you look at the portfolio of any network hardware vendor, metal is being reserved for higher-quality products, which we definitely want Omnia to join.

I hope that my brief summary makes it clear why, once we laid all the cards to the table, we have chosen a solution that is more costly for us, but which we believe will bring our customers more value. CZ.NIC is not a commercial company and we therefore don’t consider the matter from a purely commercial perspective. Our goal is to provide interested persons and enthusiasts such as ourselves with a device that will be of value, and that we can be proud of.

That’s why we set the change of the case material as a stretch goal for the important milestone of USD 500,000, at which point we were sure we could afford it. By the way, the campaign has rapidly accelerated from the moment when we announced and later reached this stretch goal. Let me thank you one more time for enabling us to reach this milestone. I firmly believe that now you understand the reasons that made us change the case and that even you who doubted the proposed solution at first will appreciate the higher value that our little hardware baby will bring to you.

Author:

Comments (9)

  1. Hans G. says:

    No complaints here

  2. Juergen says:

    I liked the design of the plastic case, wheras the metal case looks … just like a box. But I don’t care. The Turris Omnia is such an ambitous project, that I’m still happy to support it. Because inside the box there is freedom (open source software and mostly open hardware) and security (PGP-signed firmware updates and an adaptive firewall). And both means much to me in these times of lost control of our devices and of attacks against our privacy. Much more at least as nice case.

  3. Zimmie says:

    Don’t forget that a metal case can give you better segmentation of RF noise. That can be a big issue in devices with radios, and it only gets worse with more bands. The broader bands in 802.11ac makes the problem even more serious. A nice slab of metal between the logic and the antennas does wonders.

  4. Eric Marnell says:

    I’m all in favor of the metal case, for all the reasons you cite. Always consider metal to be the sign of a higher end product.

  5. John Rask says:

    I love the looks of the plastic case, and the looks of the metal case is at best boring or even worse anonymic. I accept Your argument thou; better have a great metal router for many years, than a nice looking one for a couple of years..
    I love Your entusiasm! Working for a firm like that would be really nice! (o;

  6. tyson says:

    metal creates parasitic charges, which would be a bit worse than plastic in this case. try carbon or some natural material via biomimicry to neutralize toxic frequencies

  7. Steven says:

    I assume that the metal case will be grounded. The metal case is a superior RF enclosure if done right which should be the case. I don’t care how nice the plastic case is alleged to be visually attractive. Metal enclosure done right for an RF containing instrument is vastly superior as has been well cited. Do you want a pretty ornament or a superior functional unit? Seems to me to be slam dunk, all things considered.

  8. Jarda says:

    Metal case is great but dualband antennas? Realy?
    Even if I’m not counting with situation that both cards transmiting signal and thus is possible some sort of interference is here other possible problem. It’s about customizing shape of wi-fi coverage and when I want to use directional antennas for 2.4GHz then I sacrifice part of 5GHz and possibly I’ll have problem with interference of combined (2.4 + 5 GHz) on 2.4 GHz antenna.

    Tyson: What do you mean by “toxic frequencies”? As Steven wrote, metal case should be grounded and then it creates great EM shielding for internal components and also blocks every EM signal from inside so only transmiting parts of router are antennas.

  9. Kim says:

    Shiny ugly plastic vs cool functional metal? Cant even see any arguments worth considering in favour of the former… Just a massive bunch against it… Im VERY happy you went with the sensible solution of metalchassis in the end. I couldnt care less for ‘pretty’…. Its a tool for gods sake!… Even the whole wifi-debate seems a bit strange to me but admittedly I consider wifi an emergency solution as opposed to strictly wired whenever possible.
    Oh, and a big ‘thankyou’ for setting a great project into action! Can hardly wait to get to ‘play’…

Leave a comment