Who decided to go with the small RGB LED on the Photon. Thank you!!!
This is true. It’s on the shortlist for this or next Sprint, and given how fast @mdma works, it should be pushed before you know it.
It is tricky. One could use an external uC as an reprogrammable event manager that controls the enable pin for the regulator on the Photon or control a MOSFET to control the power fed into the Photon. There are numerous nanowatt micros available that'd be able to do this.
@Carsten4207 we had a previous Elite member who was designing a power shield with an MPS430 to control charging and be able to turn off the Core (at the time) and then turn it back on based on time or event. My understanding is that the MPS430 is tiny and very low power.
The RGB LED on the Photon was actually more of a space constraint issue than a performance issue, although the new one looks great. We were deciding between about 12 different LEDs in that smaller package–the Cree, and a dozen or so Chinese clones, and only the Cree LED had adequate light diffusion. It’s way easier to get even mixing between colors in larger LEDs with large wells, so it was a bit of a challenge to identify a suitable replacement!
1 Like
Q1: Is there a way of ‘preventing’ firmware readback on Photon - similar to setting lock bits on AVRs?
Q2: What have you learned to help speed up rollout of new products like Photon? What would you do different next time?
Three related ones:
What did you do before ending up at Spark/Particle(/Sparticle)?
How did your end up there?
What do you like best at working(playing) there ;)?
1 Like
Probably stated before/elsewhere, but, as the Hardware Engineers, what are the 5 most important changes from the Corre?
1 Like
I remember that we were a little nervous making that switch since the previous LED came in an industry standard PLCC package and there were plenty of suppliers for it (including knock offs). The LED on the Photon comes from a very reputable manufacturer (CREE) and is the only source.
I feel like this blog post does a better job at capturing the differences
4 Likes
To revisit the Vbat area. We have a battery on Vbat.
Assume we are powering from a solar panel, and the 3v3 fades out, and comes back in. Should the Photon, switch back and fourth gracefully, or can we expect it to be all messed up?
Maybe not a question for the hardware guys, I am not sure.
We have learned a ton. We're actually starting to work to formalize it into a documented process that we can share with the world--a description of our version of "Agile Hardware".
Prototyping on the ground in China does a lot for us, but from a process perspective we're still learning about how to handle the measured risk of working on multiple steps of the development process (component sourcing, purchasing, packaging, testing) in parallel, rather than in series.
- I'd have to dig in and check about the read protection of flash. I know we currently lock the bootloader memory space to prevent it from being erased, but unclear off the top of my head if that's read protected as well. There is mention of read protection in backup SRAM:
RM00033 Manual
The backup SRAM is not mass erased by an tamper event. It is read protected to prevent confidential data, such as cryptographic private key, from being accessed. The backup SRAM can be erased only through the Flash interface when a protection level change from level 1 to level 0 is requested. Refer to the description of Read protection (RDP) in the Flash programming manual.
- We are always learning every day and I would suggest reading our Proto2Prod blog where we document much that knowledge for you all to consume!
I have no regrets on the Photon design process, our team is amazing and we accomplish a ton of stuff very rapidly!
The internal clock needs a stable voltage supply for it to work consistently. By enabling the BOR (brown out reset) and setting the threshold properly the Photon should be able to reset gracefully.
The new power shield should be able to make the transition from a solar cell to the battery much gracefully without resetting the Photon.
@Jack, I think it’s best if we move the VBAT discussion to a dedicated topic, or better yet, wait for the promised example usage from the nice folks at Particle.
This isn’t really the topic for it, and will probably create more confusion than it will clear up.
3 Likes
@will
tnx … will look forward to reading that. Seems like it can take at least 6 months to get new products turned round, by which time it can be legacy or have a very short lifespan! … (General comment btw, not specific to Photon)
For sure...a quick time to market is one of the best competitive advantages a company can have, and it's a skill we're trying very actively to develop 
My top six (no particular order):
- Single sided PCB, you can surface mount it down in your own design (version without headers)
- Better MCU (120MHz 1MB flash 128KB ram) vs. (72MHz 128KB flash 20KB ram)
- Better Wi-Fi (Broadcom radio) with u.FL AND chip antenna on the same board.
- Lower power consumption / Higher efficiency power supply / less heat.
- Modular firmware! Fast OTA programming now (a second or two for most apps)
- Mo' beautiful!
5 Likes
@BDub
Thank for that. Once the photon is shipped/arrives I will start consuming the various manuals & docs. Probably a simple answer, just waiting to be read 
I may not understand what this topic is about. Sorry if I ask to many questions.