Olimex STM32-P107 board ( for ethernet! )

LeafLabs Garden » Development

Olimex STM32-P107 board ( for ethernet! )

(39 posts) (4 voices)

Started 4 years ago by dreamcat4
Latest reply from gbulmer

« Previous 1 23

dreamcat4
Member

I only have a Microchip Wifi module (WiServer) which is attached to an Arduino. Its hardly a fair comparison and the Wifi chip is really slow too (1Mbit). As expected, it performs pretty slowly. Its also proven to be pretty reliable however. The Wishield TCP/IP stack is pretty custom, built up on the uip stack which has a smaller footprint than LwIP which we are considering here for 100 mbit ethernet.

WiShield (Uip) compiles into less than 32 kilobyte Flash and runs in only 2k RAM. Its also available on Maple and can't imagine the same Wishield setup creating much overhead on a Maple if you average it out. I believe performance concerns aught to be more about handling interrupts, and who gets runtime priority rather than the memory footprint, which should be acceptable since you can tune the size of LwIP buffers. Unlike our computer there's no swap disk. So your Maple programs aught to be in contention when sharing control timeslice and not be sharing memory with the stack. Please correct me if I'm wrong about that.

I am unfamiliar with STM32 interrupt system. However if you set your program code off a callback fn() on a timer, isn't that at least guaranteed to be serviced on the timeout interval? Are there other ways to de-prioritize the servicing of ethernet / RMII interface over your own program code?

Posted 4 years ago #
gbulmer
Moderator

The difference is that a UART ethernet solution doesn't let you the option of putting a TCP/IP stack onto the Maple.

I assume I must have been unclear.
I realise that the UART-based ethernet won't use the STM32F-hosted TCP/IP stack (I'd assumed that was obvious, but in retrospect I can see how I might have caused confusion).

What I mean is, when the STM32F is running a TCP/IP stack *ANYWAY* (to support 802.15.4 wireless) then there seems to be no resource benefit to using a UASRT-based ethernet instead of a local STM32F107-peripheral-based ethernet.

In this scenario:
1. I do NOT see any flash memory footprint benefit for any services which run (above the TCP/IP level) on an STM32F107 to a UART-ethernet (Put another way, the flash space for a TCP/IP stack is consumed anyway, so the UART-stack is only going to make it worse)
2. I think that the performance of reasonable code for the STM32F107-ethernet-peripheral (using DMA) is going to be in
the same ball-park as code for a UART-based ethernet solution.
3. I think there may be some advantages for some edge cases, like NTP, of a STM32F107-peripheral-based ethernet.

Clearly, I could do the experiments and measure the results of STM32F107 based solutions probably using ST code.
I'd like to know what evidence folks might have to disprove assertion 1 or 2, so it is a bigger set of test results I'd like to understand.

I can see that the time to develop UART-ethernet might be less than for an STM32F107-based ethernet.

Posted 4 years ago #
poslathian
LeafLabs

yea the NVIC on stm32 is quite flexible, with programmable priority for *most* of the 64 channels. There seems to be a bug in the hardware (seriously) that requires that the I2C peripheral run with the highest priority. You can always use high priority timers, or independent watchdog timers to guarantee some function is going to be run with a guaranteed period.

There are also these stm32-W chips, which are kind of mysteriously unavailable but we've known about them for quite a while (a couple years now). They run a bit slower, 48MHz IIRC, but have an onboard radio PHY that works on a few different protocols (xbee included I think...)

Posted 4 years ago #
gbulmer
Moderator

The STM32W implements an 802.15.4 radio.

That is the radio under a bunch of different brandnames and protocol stacks, including ZigBee, 6LowPAN, HART, and ZigBee RF4CE
(Edit: Radio Frequency for Consumer Electronics (RF4CE) merged with the ZigBee Alliance)

I have this:
http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/USER_MANUAL/CD00297658.pdf
More info at http://www.st.com/internet/evalboard/product/251361.jsp

Which has a USB to wireless dongle with an STM32W108HBU63 and a small USB 'mote' with STM32W108CBU63
http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00248316.pdf

Edit: Farnell are selling the STM32W108CBU64TR chips here:
http://uk.farnell.com/stmicroelectronics/stm32w108cbu64tr/mcu-rf-32bit-128k-flash-48vfqfpn/dp/1892383

Posted 4 years ago #
dreamcat4
Member

What I mean is, when the STM32F is running a TCP/IP stack *ANYWAY* (to support 802.15.4 wireless) then there seems to be no resource benefit to using a UASRT-based ethernet instead of a local STM32F107-peripheral-based ethernet.

Unfortunately I can't see any future announced ST MCUs which continue that integrated 802.15.4 capability. Perhaps they are dropping 802.15.4 entirely? Which is a shame but worth clarifying with an ST representative.

There isn't any obvious successor for the WiShield device yet either. Its not as easy to find a manufacturer who will sell both the module plus suitable drivers. Indeed ST seem the only obvious party who would be interested in producing an 802.11 module with the drivers appropriate for the STM32.

Posted 4 years ago #
dreamcat4
Member

Perhaps if the Marvell AVASTAR series would ever become widely available. Thats a pretty decent wifi + bluetooth module over SDIO interface. I AVASTAR since its meant to replace the Marvell 8868 which supposedly already has an open GPL wifistack.

If you can put an empty SDIO slot on the F207 board instead then that would allow the option to install a future SDIO wifi card. And perhaps without giving up the (very commonly used) SPI1 and so on. Which would be otherwise be permanently occupied when implementing an RMII interface.

And if no appropriate wifi card materializes. We can still use the SDIO slot for extra storage. Another nice thing might be an smd spot to place extra RAM.

Posted 4 years ago #
dreamcat4
Member

The Marvell chip is better known as Libertas.

http://wiki.laptop.org/go/Libertas
http://wireless.kernel.org/en/users/Drivers/libertas

Posted 4 years ago #
gbulmer
Moderator

Unfortunately I can't see any future announced ST MCUs which continue that integrated 802.15.4 capability. Perhaps they are dropping 802.15.4 entirely? Which is a shame but worth clarifying with an ST representative.

I got my STM32W development kit directly from ST Micro in the UK at the National Electronics Week Show last month.
The UK MD said distributors like Farnell would have them around about now.
They were pretty convincing that it isn't going to be dropped as a product.
The fact that Farnell are now selling the chips, but weren't a month ago, helps me believe them.

Posted 4 years ago #
gbulmer
Moderator

If money is no object, the Atmel H&D WiFi SDIO card might be a candidate:
http://store.atmel.com/PartDetail.aspx?q=p:10500186;c:100118

I got one when I went on a Atmel training course last year. Impressive, but a bit odd. Its firmware had to be loaded each time it was powered up.

I'm more interested in 802.15.4 networks, it is much lower power.

Posted 4 years ago #

RSS feed for this topic

« Previous 1 23

Reply

You must log in to post.