Dweller
Member
Hi.. while browsing for cheap STM32 boards, I found..
http://www.soliddepot.com/spruce/
Which may seem oddly familiar to people here! does appear to be a really bad search&replace rip of the docs here :( although the board itself does sound kinda interesting.. A bit more hunting finds variants without the ether, with 2 usbs.. and 16mb spi flash (for around 40$)..
Anyways.. just wondering if leaflabs had seen the webrip.. maple, oak, willow.. spruce?
robodude666
Member
At least they were nice enough to open source the software & hardware files.
Still not nice that they're using the LeafLabs logo and referencing the LeafLabs Maple in the docs.
Dweller
Member
The best bit is where here, at http://leaflabs.com/docs/maple-ide-install.html we have..
Note Note that while these instructions work on Windows XP, changes in Windows 7 mean that you won’t be able to install the IDE without disabling driver signing on your computer. We’re working on resolving this situation. For now, users on the forum have reported a workaround.
Which leads to BionicBadgers post http://forums.leaflabs.com/topic.php?id=73#post-788
And over there.. they have..
Note Note that while these instructions work on Windows XP, changes in Windows 7 mean that you won’t be able to install the IDE without disabling driver signing on your computer. We’re working on resolving this situation. For now, users on the forum have reported a workaround.
Which.. I thought would either go nowhere.. or back here. but no! they have it link to a helpful post by one of their forum users ;p..
Yep.. over at https://www.soliddepot.com/forum/viewtopic.php?f=10&t=62 .. user 'jsone' seems to have not only found the same solution, but the exact same words to use to express it ;p .. funny that.
Many more amusing bits laced throughout where they haven't quite purged the leaflabs references, or RET6 mentions..
mikep
Member
I have two of these boards because I was looking for the onboard Ethernet, larger ROM/RAM and additional GPIO pins. The thought of using the libmaple or the MapleIDE was intriguing.
They had to make modifications to the bootloader and some libmaple files due to the different CPU that was not used in any of the LeafLabs products and other subtle hardware differences. Unfortunately they just forked the LeafLabs code base and did not work with LeafLabs on a combined solution. Also the port to the CPU and hardware used on the board was not complete (e.g. there are no Digital Pins defined for the GPIOE port). I also ran into an issue where I had to patch their bootloader (http://forums.leaflabs.com/topic.php?id=1229).
Not sure why it was done that way, but it's certainly frustrating.
But now I am at a point where I can load software using MapleIDE and I got a simple webserver working.
I have not looked into getting the SDCard (which uses SDIO) or the TFT Display working.
Also, they have the onboard SPI flash using the same SPI port as the Ethernet chip, including the same CS pin. That seems wrong to me.
There are some other Maple clones out there, but at least they use the same CPU and seem to be hardware compatible.
My main issue with the Maple (I have a few of those as well, incl. a RET6) was that the onboard voltage regulators do not supply 5V to the shield on the corresponding arduino pin and they are so wimpy that you cannot really drive an ethernet shield with it.
Dweller
Member
Hmm.. the extra ram would be real handy.. I'm using an SPI FRAM at the mo to act as a buffer before pushing data down to sd-card.. was thinking I could use the greater SRAM to avoid that.. but if the sdcard is using sdio, I think I remember reading the docs for that are only available via NDA (which would kill my current sd card storage) {although.. surely there must be sample code for the cpu sdio somewhere?}
I hadn't noticed the SPI Flash & Ether shared the same pins.. I wonder if that's an error.. or if the commands are different enough not to clash? might be amusing if writing data out to ether triggered a page read from the ram ;p
Ebay had some boards sold in the past described as Stm32 spruce, that didn't have the ether, and appeared to have 2 usb instead.. (with tft & a few buttons, same cpu, and 16mb spi flash) ..
I hit the same issue with maple, mainly coming from a 5v arduino board, solved by using an external switching 5v regulator, and dropping that to 3.3v just for the maple.. gave me both supplies with enough oomph to work with.
You didn't mention in the other thread what was odd with the button wiring.. or the change required.. are you using the "solid ide" ;p or maple with your own changes? (or.. solid with your changes?)
The price is low enough to consider getting one to play with, but sounds like it might be 'entertaining' to get the basics going.
mikep
Member
Here is the hardware difference that required a bootloader change:
- both boards have a BUT button that you need to press during the bootloader initialization to get into "perpetual bootloader" mode
- on the Maple Board, the BUT button goes to +3.3V via a resistor
- on the Spruce board the BUT button goes to ground
So in the bootloader you basically have to reverse the logic to detect a button press, otherwise in my case it stayed in perpetual bootloader mode.
The change was in https://github.com/soliddigi/SOLID-STM32-Arduino-Bootloader/blob/master/main.c, Line 50:
"readPin(BUTTON_BANK,BUTTON" becomes "!readPin(BUTTON_BANK,BUTTON)".
Once I figured that out and recompiled and flashed the bootloader it was all working.
I even reported the bug to them, but they didn't get it, sent me download links to their bootloader and the IDE instead, repeatedly.
I am using the "Solid IDE" btw. But you could probably also use the regular Maple IDE if you add the same sections to the "boards" files that define the Spruce CPU and LED_PIN etc. I believe the IDE is otherwise exactly the same.
I haven't looked through all the changes in "lib spruce" yet, but you'll see changes in lib_spruce / wirish / boards where the hardware changes are addressed. Overall there should not be too many changes.
kohlrak
Member
Yeah, knowing little about arduino, i ended up buying one of these things thinking that it was actually one of leaflab's (bought it off of amazon). I've been trying to get ahold of SolidDigi all weekend to try to get support for this board, since i'm running into issues getting anything loading from the IDE. I'm having trouble getting the user code to run at all (sometimes it seems to happen by magic, but i have no idea what it is i'm doing to get it to run). It seems to upload fine, but it only runs it under special (and inconsistent conditions). I'm at a loss outside of just trying to get things to piggyback on the bootloader. I'm not sure if that's what your changes aim to fix or not.
mikep
Member
My board was behaving erratic as well. E.g. I was usually able to upload user code right after a hard-reset after plugging in the board.
- did you upload the bootloader from Soliddigi or did your board come with the bootloader?
- does the on-board LED just keep blinking?
- try holding down the BUT button on the board while uploading code from the IDE and keep holding it down until after the the bootloader is done (about 2seconds or so), see if that lets you load code from the IDE consistently.
kohlrak
Member
My board did not come with the bootloader (which was part of the confusion i had when i first got it). I eventually managed to upload the Soliddigi bootloader to get to where i'm at now.
And yes, the onboard LED (by the BUT button [marked SW1 for some odd reason, why D2 is for the LED]) blinks quickly then slowly without changing pace or going into any other modes.
And when i hold the BUT button, it does not change. Actually, it throws it into bootloader mode instead of running the code. So, i hit the reset button and it behaves normally (by normally, i mean doing nothing other than blinking).
Funny you would mention the BUT button, though, because on random occurances i've been able to get it to actually run the code by holding the BUT button while hitting the reset button. Sometimes it'd fly into "perpetual bootloader mode" and sometimes it wouldn't.
poslathian
LeafLabs
They are really shipping these things without a bootloader? tsk tsk.
Not sure why there should be too many differences with the bootloader behavior, its also possible some of their buttons are shoddy and not working 100% right electrically. On a proper Maple board things should generally just work. If they dont you can almost always just hit reset as soon as the IDE tells you its "searching for DFU device" and if all else fails there is perpetual bootloader mode.
kohlrak
Member
Ok. I got code running (it's the button problem as mentioned above, but i'm trying to write my own bootloader from scratch and i'm avoiding that button), but i'm sort of desperate to get something on that lcd. I haven't the foggiest idea how to program the lcd driver, since i can't even get the datasheet on it. I know as much s that i have to use the FSMC, but i don't even really know how to use that. Could someone make an example program, please? I just need the thing to turn on and give me some color.
mikep
Member
There is some sample code you can download:
http://www.soliddigi.com/download/showdownload.php?id=50&lang=en
It's not for the Maple platform, but just some C sample projects.
I think one or two of the samples are for the LCD, in additionnto the source code, there was a .hex or .elf file in there somewhere that you can upload to the board.
Hopefully that'll get you started.
The other thing you can look at is "CircleOS" over at stm32circle.com, it's for the "EvoPrimer" tool which also uses a STM32F103 MCU and has an LCD connected to FSMC.
Hope that helps,
Mike
kohlrak
Member
I'm gonna have to try to find a way to compile that stuff (no make file, but i gotta make sure it works before i try to use it), then try to minimalize it. Right now it's in the usual spaghetti format i'm getting used to with alot of example code these days. (i just love when i see a macro like "#define function(x) x" as it ends up being an extra step with no purpose other than keeping people from understanding the code. But i digress: i have alot of work ahead of me on this one.)
Anyway, thank you.
mikep
Member
Yes, I know, their sample code is a pain. I never got around to compiling it yet since I don't use Eclipse. I just saw that they had 2 pre compiled ones in the sample code somewhere and I uploaded those just to see if the LCD works. Hopefully I can get back to this in a few days.
You must log in to post.