You can select up to 3 cores, feel free to add comments in a post.
Would be interesting to know if some core is the most popular for F1 boards, but when it comes to F4 a different core, and when it comes to F7 a different one.
EDIT:
To clarify these are the different cores in question:
Roger’s Libmaple cores:
https://github.com/rogerclarkmelbourne/Arduino_STM32
Steve’s Libmaple cores (only F4 is diverged from Roger’s):
https://github.com/stevstrong/Arduino_STM32
Danieleff Generic:
https://github.com/danieleff/STM32GENERIC
STM’s official core:
https://github.com/stm32duino/Arduino_Core_STM32
Stm32duino and Arduino to me are mostly stm32 and have chips on s board. I use them more as chips than as stm32duino or Arduino.
I’m thinking about moving to the official STM core for the reasons that Ray has written in his last post. Maybe not for the Bluepill tho, I would miss the USB bootloader, which is quite convenient to use.
[Dust – Mon Apr 09, 2018 6:48 pm] –
I think I’m using the Generic F1 core with Bluepill.. Ok, I am, but for a newbie around here it is quite confusing to understand what is going on and this core just worked right away.I’m thinking about moving to the official STM core for the reasons that Ray has written in his last post. Maybe not for the Bluepill tho, I would miss the USB bootloader, which is quite convenient to use.
I believe the usb bootloader may be supported.
[victor_pv – Wed Apr 11, 2018 2:02 pm] –
I have to say I’m a bit surprised so far with the results for cores other than F1. Please keep voting if you haven’t voted yet.
I just voted for Steve AND Rogers F4 core, because my first F4 board is on shipment and I have no clue which core I’m going to use ![]()
Man is a creature of habit, so I would prefering to stay on leaflabs core, cause the F1 variant is nearly perfect for me.
F3 core: This is not really a suprise, isn’t it? ![]()
The bluepill with an F3 transplant as Racemaniac did is an interesting one for testing, but I just find it easier to jump on the F4.
I’m more surprised on the F4 cores, but as you said Man is a creature of habits…
Let’s see if more people keeps voting what’s the end result. So far I would say it makes sense to keep the F1 libmaple version living as long as possible, but it’s hard to say what F4 core will get the most traction.
I only added the F3 core to the repo to save it from being deleted and lost.
Someone else ported it, from an older version of LibMaple and it’s never been updated
For the F3, Avik De’s core , aka Koduino, is much better, but it seems to have gained very little traction.
I am not surprised most people use the F1.
If people want more processing power they normally use something completely different, like an ESP32 or a RPi Zero etc
If you could get a F4 board for marginally more than an F1, I am sure people would spend an extra dollar or two, but as they are 5 times the price and generally much larger boards, they are not physically suited to a lot of projects
We know that the F103 is very cheap and aggressively priced in China but the price difference is not so big. For example, it is possible, a blue pill with L476RC for 4-5$.
I don’t understand why there are no cheap boards with small F4/L4 MCUs.
Check the forum threads for bluepillF4, there is 1 about development, and another one listing some basics of what’s components to change when soldering different MCUs.
Is an open design, so the idea is to hopefully get some manufacturer in china to produce them in mass at a cheap price. That would make for a great and cheap F4 alternative to the popular bluepill.
Can you send a link to where you can get the STM32F401RB for $2.15
I can’t find them listed that cheaply.
Or was that the 10,000 or 100,000 off price ?
http://www.st.com/en/microcontrollers/stm32f401rb.html
I compared the official prices @10000, to have a common base for comparison.
This of course means that it may well prove even cheaper per unit, through local distribution channels within China itself.
the price difference is not so big.
What’s so big or not is quite subjective and likely case specific.
What does the increment 50cents or more buy you that’s not available in the 103?
[dannyf – Fri Apr 13, 2018 2:48 pm] –
the price difference is not so big.What’s so big or not is quite subjective and likely case specific.
What does the increment 50cents or more buy you that’s not available in the 103?
A faster ADC, more mips, an FPU.. wider voltage range… lower uA/MHz a USB 2.0 full-speed device/host/OTG controller with on-chip PHY, more flash and RAM, more SPIs and a bunch of other stuff. I guess you would need to compare the specs to see all the differences.
In 9 out 10 applicstions I often find myself turning down the clock speed. Obviously that it very user specific bit it should tell you that higher specs aren’t always useful.
If you can get a cortex M4, and all the extra bells and whistles for a few cents more, then why not go for it.
If on the other hand the price point is significantly higher, then there may well be better solutions.
It all depends on your needs. If you have a use for a tiny bluepill form factor USB OTG device, then this 40X may be your solution. The bluepill form factor is obviously a big draw as there are lots of other F103 boards,but none are selling in anything like these numbers. The bluepill layout seems to have hit the current sweet spot.
I have not tried the STM8, mainly because it doesn’t GCC.
BTW. I noticed some activity by ST on GitHub on ther STM8 Core, but I did not ask Frédéric what was happening. I think their original timescale for the Core was to be ready by the end of 2017 but I know they had problems and the timescale had to change.
I don’t know if the STM8 uses a Harvard or VonNeuman architecture, I should look that up, as it would be less attractive if the STM8 was the same as the AVR ATMega devices.
BTW.
How stable if the F4 Pill board know ?
Is it at a point where it would be worth getting 100 boards mass produced ?
But I haven’t tested the 32Khz crystal, mutiple pins, and the battery charger circuit.
Several more people have some assembled, but I think no one has completed all testing yet, so we may still find issues here and there, or replace/add components for convenience (such as adding sdio pullups if we pursue an F1 version, or sdio still has some trouble with F4 mcus). But probably not many changes, just a couple of corrections.
Thanks
Looks like it needs a bit longer for all the potential bugs to come to light.
I get the feeling that feeling that if a company like RobotDyn (or any company really) decided to make the board, it would probably be retailed at $10, as RobotDyn seem to charge an extra dollar for to install the bootloader onto the Blue Pill, and I doubt it costs them anything extra as often the Blue Pill etc come preinstalled with some sort of Blink sketch already.




