Easiest way to add Bluetooth communications to a Microchip PIC project

Question

We are at the tail end of a development project that is currently based on the PIC12F683 microcontroller. Fairly simple project: a couple of A/D channels, a bit-banged DC-DC converter, an output firing stage that uses IGBTs. The prototypes are working well.

Now that we are about to deliver the final design, the client has made two requests.

1. Add another A/D channel.

2. Allow field programming of the A/D threshold values using Bluetooth connectivity.

The first request is easy - simply migrate the current code to the PIC16F688.

The second request is also fairly easy - we can get serial comms using the EUSART module in the '688. That would be fine if we wanted to use wired comms (either Serial or USB).

But the client has specifically requested Bluetooth. Moreover, I am certain the customer is also going to ask us to develop apps that run on Android and iOS. They haven't asked yet but I'm certain it will come.

I've been looking at the plethora of ready-made Bluetooth modules that are currently available as well as looking at migrating the entire project over to something like the ESP32.

I'm asking for advice and the easiest and most cost-effective way to proceed.

Something I should mention is that most of our products have extremely-long lifetimes. I need to be able to source the parts our products use for 20 or 30 years from now.

Microchip makes that easy - to my knowledge, they have not deprecated any of their mainstream microcontrollers. But I can't be certain that will be the case for 3rd-party Bluetooth modules.

Advice greatly appreciated

Answer 1

ESP32 is a good choice in terms of development time. They're pretty easy to work with and there's a lot of resources out there. But you mentioned lifetime, and for long lifetimes, I'm not 100% sure I'd rely on ESP32. Even though Bluetooth will probably change considerably in the timespan of 20 years. So I wouldn't expect a product with BT to have a lifetime that long. And maybe Espressif could potentially provide those chips for so long, but I guess nobody knows.

Nevertheless, to avoid any doubts about that, I'd be looking at BT chips from manufacturers such as Texas Instruments or Nordic. They're pretty established and have had BT chip families available for a long time. So my guess is that if you chose one from the latest generation of these two manufacturers, for example the CC2642 from TI, it should still be available for quite some time. TI has excellent documentation, Nordic is worst in those regards, but there's lots of material available on the internet about the Nordic chips, so it shouldn't be very hard to get going with any of those. But it will require some development effort, that's for sure. Plus, you'd be looking at RTA certification costs to put this product on the market.

A solution with less development effort, since you're already familiar with Microchip, is the RN4678, which is a module. The interface is based on ASCII commands, Microchip documentation is decent, so it should be pretty easy to put it up and running. The other advantage here is that this module is pre-certified, so you'd save some money in terms of RTA certifications for the main markets (Europe, USA, Canada, Japan, China...). The trade-off is cost. This solution is not so small and is most probably more expensive, since you'd still need to have your host microcontroller and add this module to your design.

Unfortunately this is something where it's hard to give a definitive answer, because there's several things to consider. If I may be allowed to give an opinion, myself I'd go with the Espressif ESP32, maybe the ESP-WROOM-32, since it is pre-certified for the main markets as well, and could perform the tasks you need without having an extra host microcontroller.