Helpful ReplyHot!MHC bug in initialize

Author
electronicsgeek
New Member
  • Total Posts : 21
  • Reward points : 0
  • Joined: 2014/05/02 03:40:18
  • Location: 0
  • Status: offline
2017/12/01 11:21:24 (permalink)
0

MHC bug in initialize

PIC32MK1024MCF064
Harmony 2.04
Bug 1 : 
 
In the Pin settings : RA0 pin 13 cannot be chosen in the functions to be AN0 . 
 
Bug 2: 
 
drv_adc_static doesnot contain the initialization for Channel 5 although it is selected in the MHC .
 inaddtion : when selecting from the channel 5 analog (class-1) default or alternate input selection , ADHS_Alternate_2_CLASS2_AN6 should be CLASS1 instead.

The 2 bug are 100% correct and i have found workarounds for them , but still it was time consuming to find the problem and should be fixed in the next version .
 
bug 3 :
not necessary as bug , but you can select in the pin settings PIN 47 and PIN 48 (SOSCI and SOSCO) as output , but in reality they will not work , and the data sheet confirms that they cannot be selected as output even if no oscillator to be used . check the attached file
 
hopefully someone will find this helpful
 
regards 
post edited by electronicsgeek - 2017/12/01 13:12:06

Attached Image(s)

#1
MicroDawg
Go for a walk!?
  • Total Posts : 100
  • Reward points : 0
  • Joined: 2015/12/04 15:06:27
  • Location: Chandler, AZ
  • Status: offline
Re: MHC bug in initialize 2017/12/01 15:04:14 (permalink)
0
THANKS!
I have confirmed and logged the first 2 bugs.
#2
jtzeng
Super Member
  • Total Posts : 304
  • Reward points : 0
  • Joined: 2003/11/07 12:42:20
  • Status: offline
Re: MHC bug in initialize 2017/12/01 19:47:02 (permalink)
0
There is also a bug with SPI SCK's.
If the Pin can be used as an Input or Output then for the MK it will only configure it as an input.
This happens on SPI6 and possibly other.
This has been logged by microchip....
 
Also for SPI there is no interrupt code generated in 2.04.
 
 

John Zigrang
JTZ Engineering, Inc.
#3
electronicsgeek
New Member
  • Total Posts : 21
  • Reward points : 0
  • Joined: 2014/05/02 03:40:18
  • Location: 0
  • Status: offline
Re: MHC bug in initialize 2017/12/02 01:30:19 (permalink)
0
bug 4 :
SYS_PORT_C_ANSEL : is generated incorrectly RC2 selected Ansel is digital , generated as 1 (analog) , i had to change manually inorder to work as digital input
SYS_PORT_B_TRIS : is generated incorrectly RB9 selected as output , generated as 1 (input) also i had to change manually inorder to work.
#4
electronicsgeek
New Member
  • Total Posts : 21
  • Reward points : 0
  • Joined: 2014/05/02 03:40:18
  • Location: 0
  • Status: offline
Re: MHC bug in initialize 2017/12/02 01:30:48 (permalink)
0
bug 4 :
SYS_PORT_C_ANSEL : is generated incorrectly RC2 selected Ansel is digital , generated as 1 (analog) , i had to change manually inorder to work as digital input
SYS_PORT_B_TRIS : is generated incorrectly RB9 selected as output , generated as 1 (input) also i had to change manually inorder to work.
#5
MicroDawg
Go for a walk!?
  • Total Posts : 100
  • Reward points : 0
  • Joined: 2015/12/04 15:06:27
  • Location: Chandler, AZ
  • Status: offline
Re: MHC bug in initialize 2017/12/04 10:27:41 (permalink) ☄ Helpfulby electronicsgeek 2017/12/09 09:31:46
0
I couldn't duplicate either the RC2 or RB9 issue - so it may be fixed for next release (I am trying on latest internal)
 
#6
jcandle
Super Member
  • Total Posts : 343
  • Reward points : 0
  • Joined: 2011/09/19 22:01:53
  • Location: Rockledge, FL
  • Status: offline
Re: MHC bug in initialize 2017/12/07 06:02:23 (permalink)
0
MicroDawg, since you are testing internal, can you check WINC1500 configued with change notice IRQ?
 
2.04 omits the port suffix on the interrupt type macro.  Code won't compile without it.  Once fixed, MHC keeps suggesting t "fix" it (ie break it).
 
Thanks.
#7
matta
Senior Member
  • Total Posts : 131
  • Reward points : 0
  • Joined: 2006/11/15 14:05:02
  • Location: 0
  • Status: offline
Re: MHC bug in initialize 2018/03/27 06:48:00 (permalink)
0
What is the work around for the AN0 not being selectable?
#8
NKurzman
A Guy on the Net
  • Total Posts : 16447
  • Reward points : 0
  • Joined: 2008/01/16 19:33:48
  • Location: 0
  • Status: offline
Re: MHC bug in initialize 2018/03/27 07:30:23 (permalink)
0
in V2.04 or 2.05?
#9
matta
Senior Member
  • Total Posts : 131
  • Reward points : 0
  • Joined: 2006/11/15 14:05:02
  • Location: 0
  • Status: offline
Re: MHC bug in initialize 2018/03/27 07:38:28 (permalink)
0
I'm on V2.03 (have some projects I was trying to wrap up before making the jump), is it fixed in 2.05? It also looks like the QEI pins aren't selectable in 2.03 at least
#10
NKurzman
A Guy on the Net
  • Total Posts : 16447
  • Reward points : 0
  • Joined: 2008/01/16 19:33:48
  • Location: 0
  • Status: offline
Re: MHC bug in initialize 2018/03/27 15:29:38 (permalink)
0
If there is a bug in an old Harmony, and you need to use it: the workaround is to code the fix your self. they do not update old versions.
#11
matta
Senior Member
  • Total Posts : 131
  • Reward points : 0
  • Joined: 2006/11/15 14:05:02
  • Location: 0
  • Status: offline
Re: MHC bug in initialize 2018/04/13 14:21:23 (permalink)
0
I still can't on PIC32MK1024MCF064 in harmony 2_05_01
 
In the Pin settings : RA0 pin 13 cannot be chosen in the functions to be AN0
And the QEI pins don't seem to be mappable options either. I don't know what files are used to create the options in the pin mapper but I'm guessing they aren't editable. Having this broken is a real bummer cause the pin mapper is really the best thing about Harmony.
 
edit (cause it won't let me post a new post)
I also get
"INPUT_FUNC_C1RX not valid for enumerated type (PORTS_REMAP_INPUT_FUNCTION)"
so the CAN functionality can't really be mapped either.
 
These are the type of errors I expect when nobody has ever actually tested new code before. So I am guessing nobody at Microchip has actually ran the harmony tools against a pic32MK chip.
post edited by matta - 2018/04/16 07:49:37
#12
Johnny0099
Super Member
  • Total Posts : 146
  • Reward points : 0
  • Joined: 2015/06/20 00:33:20
  • Location: 0
  • Status: offline
Re: MHC bug in initialize 2018/04/15 03:40:38 (permalink)
0
I also experienced several issues in MHC initialization, see my previous post here:
 
http://www.microchip.com/forums/m1042894.aspx
 
I've also opened a ticket on microchip support, but the case was closed without a resolution.
#13
Jump to:
© 2018 APG vNext Commercial Version 4.5