The Second Spin? What is a ‘spin’?

Electronics Engineers use the term ‘spin’ to describe the process of having a PCB design manufactured. What’s involved in a ‘spin’, and why are they so fraught with potential points of failure?

My client and I will spend weeks, probably months, starting with many conversations discussing conceptual designs, product form-factors and PCB shapes, user interfaces (UI) and user experiences (UX), and intricate details about functionality. We may also make prototypes, and finally, I’ll design schematic diagrams, run simulations, and lay out the PCB in its intricate detail.

Finally, we’re ready to have this new design sent for manufacture. I just upload it to the PCB fabricators and assemblers, then relax with a satisfied look, right? We all wish it were that simple!

When software developers click the ‘Compile…’ button, it takes seconds or minutes until it’s ready for testing. Making a coffee usually takes more time; the process can be repeated dozens, even hundreds, of times per day. This is truly rapid iteration!

When a hardware/electronics engineer clicks the metaphorical Compile button, a flurry of tedious and error-prone work commences; the uninitiated are unaware of this sizeable lump of work/time/cost which begins when they give the go-ahead to have their new design sent for manufacturing! It’s a minefield of human failure points.

eCAD Export & Quoting

I export my electronics CAD (eCAD) design to a set of files in various industry-standard formats, and because that process isn’t always perfect, I check those exported files for correctness.

I upload these manufacturing files to one, often a few, PCB Fabricators, PCB Assemblers, and Component Procurers, and request quotes. I will check with you, my client, what is your ideal timeline for delivering PCBAs for your product, break that up into the various fabrication and assembly stages, and add padding for shipping. I’ll ask for quotes with a few sets of manufacturing-time-vs-cost options: 5 & 10 PCBs in 5, 10, & 15 days, for example. Longer processing times are usually cheaper; “expedited” (e.g. 2-day fabrication) is nose-bleed expensive.

I collate all these quotes into a spreadsheet showing per-PCB prices because quotes often come back in different quantities than what I requested for various technical reasons. They can take hours, but usually days, to respond via email. A blessed few have online portals where I can click quantities and days and see immediately the influence of these and other manufacturing options on price.

Finally I send that spreadsheet to you, my client, and you might take an hour or a day to consider it. Every day there’s a deadline hour at the vendor for order acceptance; any later, our order slips to the next day.

I will also warn you that the vendor’s number of days production time should not be trusted, because fabricators “lie” as a matter of standard business practise, because you and every other customer can pay more for faster service, and lower rate customers get pushed back in the queue. There will be no apology from the vendor when -not if- this happens to us, no matter how much you or I beg, scream, or threaten never to darken their doorstep ever again; this is just how it is.

DfM Check & Fixes

Finally, we place orders. You might be thinking I can sit back and relax now? Nope! Now the Fabricator gets to work checking my design is manufacturable!

They will put the design through their semi-automated Design-for-Manufacturing (DfM) analysis algorithms to ensure my design can actually be manufactured by them within acceptable tolerances. Even one tiny little seemingly inconsequential quirk will literally “stop the clock” on their promised manufacturing time while they wait for my response. In the meantime, their other customers in the queue are overtaking us…

Some “errors” are inconsequential and can be waived, and the process resumes, but to be fair, often, the things they find wrong are important. In this phase and the following, I check my email at least every hour so as to minimise the stopped-clock delays.

DfM violations fixed, we return to the top of the process: re-export the manufacturing files, check them again, re-upload them to the fabricator, wait for them to process and check them again, and finally commence production with the due date pushed back by however long that design-checking-and-tweaking process took.

Waiting & Worrying

The Component Procurers are also in full swing, placing orders for the components used in the design, and almost inevitably coming back to me with requests for substitutions for components not in stock. This is a whole other major topic I’ll cover in a future post.

The PCB Assembler, who puts all the components (solders, etc.) onto the bare PCB, will also review my Assembly Drawing PDF. I’ll keep them updated with the ETA for bare PCBs and/or component deliveries unless two or all three services are being performed by one vendor, in which case they’ll work it out amongst themselves.

Then we wait… …days turn into weeks… If I take time off away from home while a PCB is in-process with fabricators and assemblers, my laptop comes with me, just in case.

Bring-Up Testing

Eventually, the fabricated and assembled PCBA will arrive on my desk. However fabricators and assemblers only catch mistakes that are relevant to their domain, basically ‘mechanical things’. I, the electronics engineer, still have myriad ways to screw up the design, and they won’t be revealed until the PCBA sitting on my bench has power applied to it for the very first time!

Most mistakes are discovered during this “bring-up” testing process, where every circuit and subsystem is verified for nominally correct operation. If the design has a microcontroller, I’ll have to write test code that exercises each peripheral and I/O under the MCU’s influence.

That’s ONE spin!

Many engineers will not be as forthcoming about the many pitfalls in this process, either because of ego, or not wanting to frighten their client. Forewarned is forearmed; my client is usually the best judge of their risk and how and when to hedge those risks.

And so I name this blog The Second Spin.

In a follow-up post, I’ll discuss various show-stopper pitfalls, and the bring-up testing process, that can scuttle a spin.

Sign up for The Second Spin blog for more articles like this!

Happy spins,

Anthony.