![Photo](http://www.gravatar.com/avatar/8e5c82e352ea89f21049d6476a03311a?s=100&d=http%3A%2F%2Fforums.netduino.com%2Fpublic%2Fstyle_images%2Fmaster%2Fprofile%2Fdefault_large.png)
SPI Bug? Configuring Clock_IdleState==false Has No Effect
#1
Posted 13 March 2013 - 04:56 AM
#2
Posted 13 March 2013 - 05:04 AM
#3
Posted 13 March 2013 - 05:10 AM
Netduino Plus 2:
#4
Posted 13 March 2013 - 05:17 AM
#5
Posted 13 March 2013 - 05:41 AM
Attached Files
#6
Posted 13 March 2013 - 05:46 AM
Oops. Forgot the screen shots.
Michael
Attached Files
#7
Posted 14 March 2013 - 09:18 PM
It seems like that SPI is broken, I have issues with it my self.
#8
Posted 17 March 2013 - 01:39 AM
I am successfully using the SPI interface with two chips: the ADXL345 accelerometer and the L3G4200D gyro. These two parts differ from the LS7366R quadrature counter in that they are active HIGH clock while the LS7366R is active LOW. Chris is looking into the problem with the LS7366R.
I've attached the code I'm using for the accelerometer and gyro in case you are interested.
Michael
Attached Files
#9
Posted 23 March 2013 - 05:53 AM
Were the uploads received by Secret Labs? I haven't heard anything. While I don't necessarily expect an analysis of the uploads, it would be nice to know if they were received and someone is either looking at them or has them on a schedule to take a look.
Thanks,
Michael
#10
Posted 30 March 2013 - 06:06 PM
#11
Posted 30 March 2013 - 06:06 PM
Are you experiencing the same clock idle issue--or something else? ChrisIt seems like that SPI is broken, I have issues with it my self.
#12
Posted 01 April 2013 - 11:52 PM
Hi Michael, We have filed a bug report on this and will look into it for you. I'm wondering if there is some link between the setup/hold time and the clock state. We previously verified all four SPI clock/edge configuration options...so I'm really surprised to see this. If you invert the other settings in your code, does the clock idle get flipped the right way around? What happens if you set your setup and hold time to 0ms? We'll find this out in testing...I'm just curious in case there's a quick workaround in the meantime. Thank you very much for the repro and LOGIC screen capture, Chris
Hi Chris,
I've found the problem - very strange. The chip I was trying to talk to was bad. There was a partial short from Vdd to ground. I was powering the chip from the Netduino's 3.3v supply and the short was dragging the 3.3v down to 2.48v. What is odd is that the Netduino seemed to work just fine except for this SPI configuration problem.
I replaced the chip and now the 3.3v rail is at 3.3v. And the clock idle state is correctly following the configuration. I've got to say, this is the damnedest problem I've seen it quite a while. I suppose *something* has to be first to fail when the supply voltage goes below margins. I assume that the SPI protocol is generated in h/w and that the SPI.Configuration serves to initialize the protocol engine.
The chip is still not talking back to me correctly, but that's probably software since I haven't spend any time debugging the actual writes and reads for correct content.
Thanks for taking the bug report and my apologies if you or your staff spent to much time on it.
Michael
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users