Update: Todo Items

Hey folks! Today I would like to run through the remaining todo items before I can start shipping.

PCB

I still had some wiring organization and a couple of PI pin mappings to finalize before I could finish up the schematic for the updated PCB. These items are completed and I will be focused on the PCB today (6/17) with the goal of sending to fab by end of day.

Fab is usually pretty quick these days and I should have them in hand by the end of next week, maybe the following Monday/Tuesday if it goes a little long.

Once these come in, I will have to populate and test for a bit. It is not rare for there to be some small issues with a new PCB version, usually a forgotten or incorrectly routed trace that may need to be fixed. I do have a fairly robust planning/confirmation process to run through before sending to fab, so I’m hopeful that we will get it perfect on the first shot. But I do want to clearly state that there is a non-zero chance that we may need to do a .1 revision before it is good to go.

The primary reasons for the updated PCB as noted in previous updates are:

  • Moving control for everything except motors from the microcontroller to the PI.

  • Moving from JST to RJ45 cabling to simplify wiring.

Once these are in, wiring will be 90% plug and play, with only a small number of cable terminations to add to some hardware.

Cone Mech Mount

This is less of a mount and more of a tray, as the mech can be moved in and out with out any fasteners. The primary function here beyond holding the mech is the integrated strain gauges that measure the weight of the coffee coming out of the grinder and into the pour-over cone. This system actually weighs the entirety of the cone mech, and gets the specific grind weight by analyzing the increase from the pre grind zeroed value.

The voltage fluctuations for the strain gauges are so small, that the wires from the strain gauge to the amplifier circuit are extremely thin and delicate. I have previously designed in internal routing to protect the wiring, but this routing needs to be updated due to integrating a load combiner to simplify the load cell wiring itself.

For those that may not know, 4x load cell wiring is super super annoying as shown in this diagram:

Wiring diagram pulled from of EEVblog.

The newly integrated load combiner allows each strain gauge to be connected directly to the combiner, without having to be connected to the other strain gauges as shown above. This small change removes a ton of assembly and kit prep complexity for me and results in a much more robust circuit.

The internal routing updates required here are small, and will not require any outside fabrication. So no lead time added but it will be a few hours for me to finalize the setup.

Other Items

These other items are basically done and just require some time to bring to 100% completion:

  • Convert a couple of remaining API endpoints to use the updated pin mappings.

  • Publish the repos for the frontend and backend code bases.

  • Finish calibration and testing of grind weight code. Primary task here is feathering the grinder power a bit as we get close to the desired grind weight so it does not overshoot. The goal is to have a sane and “good enough” value here, not perfection. This value can be improved as we get more data from the brew process itself.

  • Run through the frontend and sanity check functionality. I currently do not know of any missing functionality, but it would be normal to find a couple of small bugs. I will share results of this run down next weekend. Anything that does not break functionality will not block completion or release.

  • Set up a calibration/tuning script for the water heater. Due to the nature of a heating elements manufacturing process, there may be some small differences in heating time across different heating elements. This initial calibration will make sure the code is properly tuned to its machine’s heating element.

  • Cloud Control going live.

This may not be an exhaustive list and I will update everyone if any items are added.

Shipping

Still no hard date but it’s close. I think a fair and realistic estimate is 4 weeks. This gives me two weeks for the allocated tasks with a nice buffer for other things that pop up. My goal is to be shipping by Monday, the 17th of July

Any future updates will include remaining tasks and how likely we are to hit the July 17th goal.

Quick Travel

I will be traveling from Sunday to Tuesday. Will be back to Mugsy on Wednesday.

Thanks and I hope everyone enjoys the weekend. See you soon!

MComment