OK, trying now, BRB

Switching to serial and then back again makes no difference.

I cannot install the MFUSB.ini because windows says:

The specified location does not contain information about your hardware.

With only OK as an option.

So I guess were stuffed then.

The only thing I thought about was the bootloader being at fault.

OK, trying now, BRB

Hi DrJaymz,


Can you please try two things really quickly?

1. Open up your project properties and go to the ".NET Micro Framework" tab. Change transport from USB to Serial. Then change it back to USB. Try to run your app again.
2. If that doesn't work, please try switching back to the MFUSB drivers.

If all that fails, there are two final options:
1. If you're running Visual Studio 2012 (or would like to try Visual Studio Express for Windows Desktop), upgrade to the .NET MF 4.3 beta SDK. It has dramatically better diagnostics when deploying and it supports .NET MF 4.2 boards as usual.
2. Reflash your Netduino with the production 4.1 firmware. That's really a last resort, but if Visual Studio is failing to connect...it may b a bug which needs to be fixed in the NETMF 4.3 SDK.

Chris

P.S. If you're running inside a virtual machine, the standard WinUSB drivers may not be set up to handle that. Switching to the MFUSB drivers may be a solution there.

Hi DrJaymz,

1. Install the .NET MF 4.2 QFE2 SDK
2. Install the Netduino 4.2 SDK (latest version)
3. Upgrade your board to the latest Netduino firmware

Please let me know how it works for you,

Chris

Done.

It will not deploy now. Whatever I try to do.

Incrementally deploying assemblies to device
Deploying assemblies for a total size of 1456 bytes
Assemblies successfully deployed to device.

You'd think its deployed then? No. The error comes up:

Error 1 Device not found or cannot be opened - USB:Netduino

Using MFDeploy I can get

HalSystemInfo.halVersion: 4.2.0.0
HalSystemInfo.halVendorInfo: Netduino Plus (v4.2.0.1) by Secret Labs LLC
HalSystemInfo.oemCode: 34
HalSystemInfo.modelCode: 177
HalSystemInfo.skuCode: 4097
HalSystemInfo.moduleSerialNumber: 00000000000000000000000000000000
HalSystemInfo.systemSerialNumber: 0000000000000000
ClrInfo.clrVersion: 4.2.0.0
ClrInfo.clrVendorInfo: Netduino Plus (v4.2.0.1) by Secret Labs LLC
ClrInfo.targetFrameworkVersion: 4.2.0.0
SolutionReleaseInfo.solutionVersion: 4.2.0.0
SolutionReleaseInfo.solutionVendorInfo: Netduino Plus (v4.2.0.1) by Secret Labs LLC
SoftwareVersion.BuildDate: Sep 19 2012
SoftwareVersion.CompilerVersion: 410894
FloatingPoint: True
SourceLevelDebugging: True
ThreadCreateEx: True
LCD.Width: 0
LCD.Height: 0
LCD.BitsPerPixel: 0
AppDomains: True
ExceptionFilters: True
IncrementalDeployment: True
SoftReboot: True
Profiling: False
ProfilingAllocations: False
ProfilingCalls: False
IsUnknown: False

So presumably the device is working. The output is set to deploy to netduino over USB.

So surprised it still can't really be used seriously.

Hi all, I have dusted off the netduino VMware system and aquired another netduino plus. I am confused because I cannot find a definitive guide to what I should be using with versions, the front page mentions 4.1 4.2 sdk, 4.2 firmware, 4.1.0 sdk for the plus and I can't figure out what I need. I have 4.2.0.0 RC4 firmware in my netduino and 4.2 SDK and latest netduino DSK but I cannot use the PWM because its all messed up and the checksums don't match and its all confused. So. Lets start over... Uninstall SDK's , Netduino and just have VS2010. Now, what to install and in what order? I'm happy to update the netduino firmware. Presumably even now, the bug where it cannot be reliably deployed is still with us?

It does deploy the application, then afterwards it says preparing to deploy and then that times out.

Switching to serial and then back again makes no difference.

I cannot install the MFUSB.ini because windows says:

The specified location does not contain information about your hardware.

With only OK as an option.

So I guess were stuffed then.

The only thing I thought about was the bootloader being at fault.

Hi DrJaymz,


Be sure that you're selecting the MFUSB_Netduino.inf file, not MFUSB.ini.

We're basically overriding the Windows driver system by forcing it to use MFUSB instead of WinUSB. Unless we select "no, let me choose" at every step it will try to use the driver specified for the VID/PID in the registry.

Chris

Yes, understand, but it will not let me. I'm running XP 32 bit.

Yes, understand, but it will not let me. I'm running XP 32 bit.

I clicked uninstall driver

Then let it install it again.

then just now it worked.

There must be something iffy with the driver caching or something - but I rebooted the VM and changed my LED blink rate just to check, rebuild and deploy - it works fine. Now back to my original issue - I couldn't work out the PWM syntax as usual the MIcrosoft help file says exactly what the function name is and that it has parameters such as duty ratio, but doesn't bother to tell you what units they might be. Really I want to start a PWM signal 10k bananas with a duty ratio of 50 pears. On say D5. Just to add to the confusion, I saw a number of questions in the forum about this and they were using the secret labs definition of PWM and were told not to use the SPOT ones. If people were using the SPOT function then PWM_0 doesn't relate to any pin so I never did get a straight answer to this.

You could always argue that the netduino is a really big CPLD. I guess it depends on what you are trying to do and quite a lot can be implemented in software and I never cease to be amazed how ingenious some peoples solutions are. I guess most uses for the netduino are not concerned with outright speed but (in my case) out of lazyness. happy coding.

Yes Thats the solution, thank you Dr Jaymz.

Yay.... I am so used to using String.Format that I really miss it, but obviously handling strings on a puny system like this is always going to be slow or memory intensive. I'm glad that F2 works for you and I don't know why I can't find it documented anywhere.

Try ToString("#.##")

ToString("F2")

??

As mysteriously as it stopped working, it started again. So, please ignore.

Yeah, it does that. A lot.

I am using NetworkChange.NetworkAvailabilityChanged to detect if it is available, the callback contains if it is or not connected, but there is no way I can see to determine it on startup. i.e. if I pull the cable in or out I can tell the status, but on startup I can't tell. Anyone else got a solution?

Has anyone got this working yet? I am using a Melexis MLX90614 IR Thermometer and I cannot get it to respond to a read from 0x07. I don't understand what the internal address register and size fields mean in this context. I have other devices on the I2C bus responding correctly, and I have pullups etc. I just cannot get a bean out of the MLX90614. I'm running 4.2.0.0 RC3. With the other devices I wrote to the device telling it what I wanted then do a read. Can I do only a read transaction? I'm about to give up on this.

I'm using a serial qVGA graphic display that draws about 200+mA, the Netduino can draw a couple of hundred as well, sooner or later I often wonder if my issues with stuck deployment and general shakeyness could be lack of power. Also my project behaves differently when running from a USB charger as opposed to connected to a PC. The timings appear to change on the UART which should be fixed surely or else it wouldn't work properly.

Yes, pretty sure that could be a power issue.



I would not be comfortable with a constant 800mA draw, even if its not shutting down, I don't like my board to be getting so warm. This could cause the board to get quite warm. I like to power via 5V rail with either a PC power supply, or a similar stand alone supply, and only use the on-board regulator for small projects.

It'll be fine, you worry too much, what could possibly go wrong....... (don't know why they worry at work when I say that)

Hi DrJaymz,

Does this help?
http://forums.netdui...-5v-header-pin/

Well it states that the 5v regulator on board is good to 800mA we know that the controller uses about 250 so that leaves 550mA to stay in the current spec. So as long as the on board reg doesn't overheat we're good. The reg will shutdown in the event it gets too hot so it should be fine. perhaps my issues could be insufficient supply decoupling.

You don't need to be concerned with the debug statements, if no debugger is attached they do nothing. With no debugger attached watch out for much faster execution which may change how your program executes if you have not been careful with delays. I had been using I2C and effectively closed it immediately after sending a shut down, in debug this worked well, with no debugger it shut the port down before it had chance to send shut down!

Well my master piece is almost finished. Its got pressure transducers, accelerometer, thermopiles and LCD touch screen. Its a piece of avionics. Its awesome. Now, I have accelerometer, thermopile and pressure on a single I2C with some low-pass filtering. This all runs on a thread. My routines for the display use the UART and they also run on their own thread. Then my main execution runs on the main thread. I have logging on yet another thread. And I'm currently using about 38kb of program space. I was wondering if there was any way to determine the CPU load - I mean all those threads have thread.sleep in them at some point and if there is a speed limit my code hasn't reached it limited mainly by the speed of the display and the I2C devices. I was thinking of sticking GPS for ground speed and heading information on the spare com port. This will require constant baby sitting of the NMEA sentences. The question is will the little netduino cope? How to tell?

I am using D4 and D5 with an twiddly knob (encoder). The common is connected to ground and the weak pullups are enabled. Thus when I the encoder connects D4 and Common I get a 0, and same for D5. I have built a state machine and configured it such that 4 - 1 represents a clockwise detent and 1 - 4 is anticlockwise. All good so far. I am using an interrupt on edge for both and when either is interrupted I check both and compare it with my state machine. The problem I am having is that I am not receiving all interrupts. The main weird thing is that it works better in one direction than the other. The speed of the interrupts is not very high perhaps 10 per second or so. If I disconnect my encoder and use push buttons I can see what happens. Holding down Button A first prevents Button B from generating an interrupt. B on its own generates an interrupt. Holding down B doesn't interfere with A. Its nothing to do with A not finishing the interrupt routine, everything ought to be symmetric. For now I have bodged it knowing that 3 - 1 is a clockwise rotation, though its not a valid set of states. I have added 10nF on D4 and D5 to ground, based on the premise that internal pull-up is about 100k, this should filter out glitches and still pass the fundamental that I want. Any ideas why the interrupts are not working as they should.

Hello all,

Just wondering if there is any eta on the 4.2 firmware? It's a little frustrating the number of NotSupported exceptions that are in the production release. I've tried flashing RC3 a couple of times now but I can't seem to get it to work, always results in an error to the effect of 'cant find entry point' after flash with MFDeploy and VS just bombs out with CNC_ENTRY_POINT_SOMETHING when trying to deploy. anyone happen to know a fix?

I'm hoping RC3 is going to be just around the corner as it will save me a load of coding for even a basic network device, but looking at some of the open issues that seem's unlikely

Thanks,
Jamie

There doesn't appear to be any roadmap for this development so I guess you are stuck with 4.1.1 for now. As for your error, did you fully erase the firmware? It might be that the bootloader is pointing to the wrong address and thus it cannot execute your program.

I tried this update to fix my issue with the I2C and repeated start. It did not fix my repeated start issue and I noticed the following.

• system.math.pow(x,y) is not implemented, which is a pain because I cannot easily implement this another way.
• (float).tostring(f2) gives a result like 1.0000 regardless of the argument placed in the tostring.
• port.read is working in an inverted manner for some reason.

Am a bit peeved to be honest.

Hi DrJaymz,

With the .NET MF 4.2 beta firmware, there are a few known bugs. The new emulated math library has a few missing or mis-calculating routines. There are a few networking and driver glitches. These are all being worked on, to make the production 4.2 firmware as high-quality as possible.

If you're looking for stable firmware and repeated-start-bit, perhaps you can try the v4.1.1 beta firmware (which is more or less release quality--we just haven't stamped it as such)? We have created an I2C wrapper which will let you do repeated start bit with it as well.

Sorry for the troubles with the beta firmware,

Chris

Thanks for your reply.

Well OK, I got miffed with it and went to the pub instead. I have posted the problems I had with it anyway so that someone can fix them in due course.

If I use the 4.1.1 firmware where is the *current* beta release and can I keep my 4.2 SDK and change the target to 4.1 in VS? If this is the case how do I use the repeated start functionality out of the available functions - is there an example of this? I guess I was lucky that all my other I2C went so smoothly but it means I didn't learn anything! So far I have not had a bean out of the Melexis Temp Sensor.

On a positive note, with the shipped firmware, I frequently got a lock up when trying to debug some code. It would get stuck on the deployment part. I would then force it with a remove and reinsert which would then make VS go nuts with a null reference error and then another error (hardware), following that it worked. When I updated to 4.2 this lock up did not occur once, but deployment was a little slower.

Before I get excited does this: 1) fix the usb deployment that fails and blue screens the machine. 2) contain the missing math functions like math.pow(x,y) which previously threw a wobbly with not implemented exception.

Upgraded my code to 4.2 RC4 no operational problems noted with execution. The missing FP operations seem back and the values are correct. Deployment still fails after a couple of debug executions, the problem is a bug in the bootloader on the netduino not with the driver on the PC. Once the deployment fails, there is no way to gain control of VS. If you remove the USB the machine blue screens. End task then remove and reconnect works. This is the bug nobody seems able to fix still after all this time? I know this because it still in the stuck state after removing the usb connection and reconnecting without resetting the board.