DfM Case Study: Getting a pricing baseline with Macrofab
Parker Dillman discusses how to optimize costs for Macrofab’s assembly processes.
We first talked with Thomas about his experience creating a low-volume product run. He was an enthusiastic user of Macrofab, an end-to-end service for creating circuit boards and box builds. The KBox is delivered fully assembled from Macrofab before being listed on Tindie and shipped to paying customers.
Today, I talked with Parker Dillman, Co-founder and Lead EE at Macrofab about some of the pricing considerations when using such a service:
Like any manufacturing service, the greatest cost savings will come from optimizing for a manufacturer’s specific processes. If we hope to understand how to improve the cost profiles and the manufacturability of this device, we need to understand how it’s priced currently. For a low-run device like the KBox, it’s entirely possible that the current price is difficult to improve without significant design changes (which may be untenable). Without understanding which elements contribute to the Cost of Goods Sold (COGS), then we can’t attempt to improve the costs and manufacturability.
Understanding the process costs
For Macrofab, we learn that the process does not actually optimize for things like placing components on only one side of the board. This is because their process optimizes by bundling together multiple customer orders and running them through reflow. Since a single board having components on both sides will impact the entire run of bundled designs, it’s not worthwhile optimizing for single sided assembly. It could be a prudent DfM consideration for other assembly houses, however.
Parker also discusses the cost adders from using a non-standard board specifications. Moving to a 6 layer board or a thinner profile board will have a significant impact on the costs of the low-run PCB. You can try this out by signing in and playing with the specifications for the PCB on the demo project on Macrofab (or view Parker doing this with the MacroDuino around the 12 minute mark of this video). Also interesting, he points out that the lead time is tied directly to the PCB specifications. We have discussed component lead times in the past and the contributing factors:
Component lead times exist for a reason
If you want to make a brand new set of silicon chips, start with a pile of sand.
The greatest savings for the KBox will come from shrinking the board size down, reducing the overall cost of the assembly. Simplifying the design and sourcing lower cost components, like we discussed when reviewing the Bill of Material (BOM). This will also significantly lower the costs which go directly to the bottom line of this device’s cost.
In future versions of DfM Case Studies we’ll take a look at simplifying and reducing the cost of the mechanical enclosure, as this was another significant cost adder. We’ll also take a look at transferring this design to other assembly houses and getting approximate costs for comparison. The low-volume nature of this device may make the initial pricing difficult, but we know this is a consideration for many of our readers.
What other DfM considerations would you like to see? Let us know in the comments!