Hello once more!

We have some good news and some bad news in this month’s update. The bad news is that we’re currently working through a production issue involving the finish of our case. The TL;DR is that we’re moving back to a frosted finish and delaying our delivery date to November. Read on for all the gory details.

The good news is that, aside from this issue with the case, the rest of our production could not be going better.

Case Manufacturing Update

Whilst preparing for our production manufacturing run, our CNC manufacturer discovered a discrepancy with the samples we previously received. It turns out that the clear transparent case we showed off last update was not produced using the finishing method we asked for, and was instead finished with a more expensive vapour polishing method.

This error was due to an internal communications issue on the part of our CNC manufacturer, and we’re working with them on a path forward. (As I was writing this update, new samples with the correct finish showed up at my door!)

The consequences of this discovery are two-fold:

1. We will be producing our production cases in a frosted finish, as we were expecting before last month's update. We apologise for the back and forth on this!
2. Our delivery date for our first batch of units will be delayed. We'll talk more about this later in the update.

Frosted finish iteration

Last month, when evaluating whether to proceed with a frosted or clear case finish, one factor we considered was that we weren’t totally happy with the result of the bead blasted samples we received. As we are now once again pursuing a bead blasted finish, we are now also working on iterating on the frosted finish to get a result we’re happy with.

The sample that just arrived has a sanded (for real this time!) finish, which looks and feels more like what we’re aiming for:

From top to bottom: vapour polished, sanded, and bead blasted with a coarse grit

We’re also investigating a finer grit of bead blasting, which we’re expecting will look quite similar to the sanded finish, but have a nicer feel in the hand.

Delivery date changes

Our original delivery date estimate for our first production batch was mid-September. This is no longer a realistic estimate.

Before this latest production hurdle, we were expecting to have to push back our shipping estimate by about a month. This was because of the delays with finalising our production case design. This new issue with the case finish has hit us in an area where all of our slack has already been used up, and so we will need to adjust our shipping date even further.

At this point, we are expecting to ship our first batch of units sometime in November. The second batch is only minimally affected; we’re expecting to be able to produce and ship those by February.

We apologise for this delay. We put enough slack into our production estimates to weather production setbacks, but we were not prepared for the same stream of work to encounter multiple issues back-to-back like this.

On the bright side, we are reasonably confident that our delivery date won’t need to change a second time. At this stage, we have production quantities of every part of Tangara except for the case on-hand already. This means that we’re unblocked with testing, flashing, and sub-assembling some parts of each unit in advance.

Production PCBs and sponsorship

Thankfully, we also have some very good news to share this month! PCBWay — our production PCBA manufacturer — reached out to us to offer a small sponsorship for our first production batch of Tangara PCBs. PCBWay offered us a discount of just under USD$1000 in exchange for a few words about their service. This seemed like a good deal to us, since we like talking about our manufacturing experiences anyway.

Manufacturing this batch actually finished very shortly after last month’s update, so we’ve already received the boards and completed initial testing of a few from the batch (with our finished testing jig!)

Our first production batch of PCBs is in these boxes!

A pair of production boards in our testing jig. The little green light means the boards are good!

Tangara’s mainboards especially are reasonably difficult to manufacture; they’re a two-sided SMT assembly with an unusual board outline, protruding components, and a small tolerance to fit within our cases. So far in our testing we’ve encountered a single unit with an apparently faulty battery management IC, but we’re yet to see any other issues (and we’ve certainly not encountered any soldering faults).

When preparing for this batch, we also briefly considered using a different colour of soldermask to the classic green. It was actually a PCBWay customer service rep who flagged that a change in colour would have an impact on the minimum soldermask width — which in turn could have led to soldering issues on some of our very fine-pitched QFN components.

That’s all for now!

Thanks for reading another exciting Tangara project update! I apologise again for the bad news this month; with any luck, it’ll be smooth sailing from here.