Taking what you said into mind I went back and tried using 10600Khz instead of 10000 just for fun.
It worked beautifully. Doubled the transfer speed and now my project is far, far more responsive. Thanks!
Sukasa's Content
There have been 24 items by Sukasa (Search limited from 27-April 23)
#63553 SPI Performance issues
Posted by
on 16 July 2015 - 03:41 AM
in
Netduino Plus 2 (and Netduino Plus 1)
#63551 SPI Performance issues
Posted by
on 16 July 2015 - 03:02 AM
in
Netduino Plus 2 (and Netduino Plus 1)
I'm on a stock N+2, so 168Mhz if I recall
#63549 SPI Performance issues
Posted by
on 16 July 2015 - 02:34 AM
in
Netduino Plus 2 (and Netduino Plus 1)
I'm working with a project and trying to write data to an LCD. It's rated up to 10Mhz speed and I've tried to set this in my code, however I'm getting far, far slower data transfer rates than I should. Being as the code is rather complex, I'll link the project github page instead of trying to give you a useless look at it.
My question is, what could I look at as a way to improve my SPI transfer performance into the LCD? I use lower-frequency devices on the bus (rated up to 1Mhz), as well as reading from the SD card at several points during the transfer. Could any of this prevent the 10Mhz transfers from running at 10Mhz?
The project code is available at https://github.com/S...Oven-Controller for reference if needed
#63033 FileStream memory requirements
Posted by
on 03 June 2015 - 02:41 PM
in
Netduino Plus 2 (and Netduino Plus 1)
Hi,
I'm working on a project that has generally left me about ~23KBytes of memory free at any given time. Unfortunately, this seems to be just low enough that at times I will get an unspecified IOException when trying to call File.OpenRead(). Roughly how much memory does File.OpenRead() need in order to successfully open a file, and is there an alternate method that uses less memory? I am opening the file in order to stream it into memory and then immediately out over SPI
#63002 Phantom SPI Clock Pulses
Posted by
on 02 June 2015 - 03:04 AM
in
Netduino Plus 2 (and Netduino Plus 1)
(copy-paste from the breakdown on the Adafruit forums)
I was able to figure it out!
The thermocouple modules you sell have 10k pullups on the inputs to their SClk lines... and the protection diodes are used to pull the clock line low. Of course, this means that at any given time these modules are backfeeding 3v3 onto the SPI Bus' SClk. Whoops!
The solution I had was to place a PNP transistor in, to make the High-Low signal from the Netduino into an Open-Drain signal from the PNP. When the Netduino pushed high, the PNP would stop conducting and the thermocouple modules would pull themselves high. When the netduino pulled low, the PNP would conduct and pull the thermocouples low. When the uC was floating, the PNP would prevent backfeeding pulses that messed up the clock signal to the LCD.
So, long story short: the Adafruit thermocouple breakouts like to pull up the SPI bus' clock line, and the Adafruit 2.2" Lcd is susceptible to that, leading to "phantom" clocks and a few nights of headaches.
#62984 Phantom SPI Clock Pulses
Posted by
on 31 May 2015 - 08:36 AM
in
Netduino Plus 2 (and Netduino Plus 1)
I've cross-posted this on the Adafruit support forums as well, but I'll ask here in case it's something on the uC...
I have a Netduino Plus 2 with two MAX31855 thermocouple breakout boards and a 2.2" colour LCD breakout, all from Adafruit. The LCD works just peachy on its own. However, if the SClk pins for either of the Max31855 boards are connected, then it seems that one or two spurious clock pulses are generated when trying to read from the LCD, one extra clock pulse per SPI read/write call (I use 2 calls, as there is an 8-bit 'dummy' read needed), which seem to happen *before* data is read in from the LCD.
These extra clock pulses do not happen if the LCD is the only device connected to the SPI bus' SClk line. Is there any reason the N+2 would/could produce extra clock pulses when additional devices are added on to the SPI bus, that I could look for in my circuit?
#62785 Cannot find any entrypoint! - The program '[16] Micro Framework appli...
Posted by
on 22 May 2015 - 02:56 AM
in
Netduino Plus 2 (and Netduino Plus 1)
This might be silly, but what are your build configurations? I commonly get this error when either no deploy checkboxes are filled in, or more than one is.
#62431 N+2 Reflow Oven WIP
Posted by
on 05 May 2015 - 01:37 AM
in
Project Showcase
Since Chris said that he'd love to see more stuff being done with netduino plusses... thought I'd share my N+2 build.
There's a fairly large imgur album already put together, so for brevity I'm just going to link that for a lot of the build itself.
Right now I'm in the last phase of development, which is just code. I'm working on the user interface for the LCD screen and getting that working, after which I'll start working on the web GUI for configuration and such. A lot of the code is written, it's just the glue logic to hold it all together, and most of the user interface, that needs to be completed.
It's a little outdated (I have a few updates yet to push from my laptop), but I keep the source code available on GIthub
There are a few people who deserve major props for this build. First up, obviously, are Chris and the rest of the Secret Labs gang who produced the Netduino Plus 2. Great little microcontroller. After that, Fredrik Rinsén deserves applause for his MVC-like web framework. I use an I2C bus library written by phantomtypist, which was derived from a version by Pavel Bansky. Lastly, I use Inquisitor's CPU load estimator to determine CPU load for the web interface.
#62378 Introducing Netduino 3 Wi-Fi
Posted by
on 02 May 2015 - 02:00 AM
in
Netduino 3
Can SDHC support get put on that feature wishlist? It's getting harder and harder to get ahold of <=2GB cards these days
#62356 Introducing Netduino 3 Wi-Fi
Posted by
on 01 May 2015 - 07:05 PM
in
Netduino 3
That's pretty awesome. You kind of missed the most important thing, though - you've brought back official +12VIN support! It's a small thing, but can be pretty important. My existing N+2 project is stuck on an 8vdc adaptor to power the n+2 as well as a pair of +12V cooling fans, and being able to upgrade to +12vdc in future if I swap uCs is a great thing.
#62350 Introducing Netduino 3 Wi-Fi
Posted by
on 01 May 2015 - 05:27 PM
in
Netduino 3
I'm also curious about the 'new' power subsystem and power management - what does that entail?
#62342 Introducing Netduino 3 Wi-Fi
Posted by
on 01 May 2015 - 04:47 AM
in
Netduino 3
SD is great bulk storage, but something akin to the 'Simple Storage' mentioned in another thread, even if it was just a few bytes would be brilliant, especially for config information.
OTA firmware updates would be pretty sweet, especially if there was a way to balance out the firmware upgrade with the managed code running and then let managed code decide when to reboot into the new firmware... But the killer feature of that would be the ability to roll back if something went wrong.
#62339 New network chip driver: Asix AX88796C
Posted by
on 01 May 2015 - 04:16 AM
in
Netduino.IP Technical Preview
Wake-On-Lan... will that factor into the capabilities of the N+3 natively, or is that not likely?
#62338 Introducing Netduino 3 Wi-Fi
Posted by
on 01 May 2015 - 04:12 AM
in
Netduino 3
Definitely the SSL and AMQP. The Wi-Fi is nice, but I prefer hardwired ethernet. The GoBus is a plus that'll only get bigger over time, too.
The extra code space feels kind of meaningless, however. Even on my N+1 I never came close to running out of room for binaries; it was always the RAM that was constricting.
#62332 Introducing Netduino 3 Wi-Fi
Posted by
on 01 May 2015 - 01:43 AM
in
Netduino 3
Wow, awesome... Will PoE ever be in the cards for a Netduino Plus?
#62314 Scott Hanselman just flashed a Netduino Wifi at BUILD
Posted by
on 30 April 2015 - 06:39 PM
in
Netduino 3
Given the prior announcement of upcoming GoBus modules, and that the N3W has the ports... I wouldn't miss the Go if the advantages of it get rolled into the N3W.
#62294 Scott Hanselman just flashed a Netduino Wifi at BUILD
Posted by
on 30 April 2015 - 03:05 AM
in
Netduino 3
So will the cost for a netduino up here in Canada be any more reasonable going forwards? I just had to pay nearly $90 to get ahold of a replacement N+2 after smoking my last one, and if that's the kind of price premium I'm looking at (and this is WITH a discount), then the price premium of a Netduino over anything else is getting to be simply too much to be worth it.
#62293 Configure TIMER to count pulses.
Posted by
on 30 April 2015 - 01:22 AM
in
Netduino Plus 2 (and Netduino Plus 1)
How fast do the pulses come? If they're more than a couple Hertz, a more reliable idea might be to use a 555 timer set up as monostable in conjunction with a low-pass filter and then read the result using an analog input.
#62097 TristatePort bug on Netduino Plus 2
Posted by
on 13 April 2015 - 11:10 PM
in
Netduino Plus 2 (and Netduino Plus 1)
Quick update to this - once .NETMF 4.4 is released, this will not be an issue. The pull request to remove the managed code crash just got merged, and setting this property to its current value will simply do nothing, without throwing an exception.
#62006 Netduino Plus 2 Firmware v4.3.1
Posted by
on 01 April 2015 - 12:27 AM
in
Netduino Plus 2 (and Netduino Plus 1)
Seems after upgrading I can no longer debug on the device; it loads the binaries onto the unit and then starts debugging, and before the entry point is hit it kicks back out to non-debug mode in VS.
VS 2013 Community Edition, update 4.
E: Nevermind this. Another project was set to deploy and because of that *nothing* deployed. I found this out by rebooting the CLR via MFDeploy, wherein it complained it had no entry point.
Debugging is really, really slow now though. And the debugger really hates attaching after deploying assemblies.
E2: After spending a little time on this, I'm noticing that my N+2, when the debugger fails to re-attach after deploying binaries, generally gets stuck in a non-running state. The LED is solid on (my code blinks it), but the device is fully responsive and I can reboot the CLR from MFDeploy at which point it seems to run fine.
#61991 TristatePort bug on Netduino Plus 2
Posted by
on 30 March 2015 - 09:14 PM
in
Netduino Plus 2 (and Netduino Plus 1)
It's a bit of a bump, but I made an issue + pull request for this on the .NETMF github and any support would be appreciated!
E: Link
#61990 MVC web framework like web server experience on Netduino
Posted by
on 30 March 2015 - 08:11 PM
in
Project Showcase
This is pretty awesome, I think I might have to make use of this in a reflow oven project I'm working on...
#61989 Netbios
Posted by
on 30 March 2015 - 08:03 PM
in
Netduino Plus 2 (and Netduino Plus 1)
There was an old thread talking about NetBIOS and getting it built into firmware 4.2.1. Was that included, or are there still a number of steps I would need to follow to get it working on my NPlus2?
#59301 SPI Help
Posted by
on 19 July 2014 - 08:05 PM
in
Netduino Plus 2 (and Netduino Plus 1)
I'm trying to work with an SPI-like device (it's a series of shift registers controlling LEDs and reading some buttons in a premade package), using my N+2.
The device has a memory map of 12 bytes light data, and 5 bytes button data, arranged linearly along a single buffer - as I write in the light data, the first five bytes out are the buttons and then the last 7 bytes out are the first 7 bytes from the old light data.
I'm trying to do an SPIBus.ReadWrite() to read the buttons and the lights at the same time. A simple write() works fine - it pushes the garbage button data and then pushes in the correct light data, so the LEDs work properly. However, trying to ReadWrite() results in garbage data getting written to the LEDs.
My code is the following:
public void Refresh()
{
int Count = 17 * _Modules;
//SPIBus.WriteRead(_Buffer, 5, 12, _Buffer, 0, 5, 0); // Doesn't Work
SPIBus.Write(_Buffer); // Works, but no input reading
_Latch.Write(true);
_Latch.Write(false);
}
_Buffer is arranged so that bytes 0..4 are the button data to be read in, and 5..12 are the 7 bytes of LED data.
What do I need to change in the WriteRead() so that it can properly write out the LED data *and* read in the button data? I'm on firmware 4.2, and Visual Studio 2010
