AnsweredHot!PICKit 4 - marketing vs. development philosophy

Author
CDS
New Member
  • Total Posts : 12
  • Reward points : 0
  • Joined: 2018/09/27 22:20:44
  • Location: 0
  • Status: offline
2018/10/08 23:54:10 (permalink)
0

PICKit 4 - marketing vs. development philosophy

Microchip's PICmicro family with it's thousands of members and it's software and development tools is a whole universe on it's own. The just as powerful AVR family is in another galaxy. Most of us decided for one world or the other long ago. The fact that this totally different cores are both owned now by the same company does'nt mean that developers will switch back and forth across the sky for their future projects. And even if they did, it would'nt create any additional sales.
 
For PICKit 4 i was hoping for a faster Pickit 3 that supports every current PIC family member with 3 breakpoints and maybe with 2 logical probe inputs on the additional pins 7+8. That way, i'd allmost get back my old ICE 2000.
Any features to support ATMEL controllers are of no value at all to me.
 
In my opinion that could have been the scope of a pretty similar but separate AVRKit device.
 
post edited by CDS - 2018/10/09 15:30:42
#1
NKurzman
A Guy on the Net
  • Total Posts : 16442
  • Reward points : 0
  • Joined: 2008/01/16 19:33:48
  • Location: 0
  • Status: offline
Re: PICKit 4 - marketing vs development philosophy 2018/10/09 05:43:59 (permalink)
5 (2)
The Hardware breakpoints are handled it the PIC itself. The debugger can not add any more. Probes synced with the execution would fall in the same category.
The ICE2000 was a real emulator. But those are all history now.
#2
Jim Nickerson
User 452
  • Total Posts : 5280
  • Reward points : 0
  • Joined: 2003/11/07 12:35:10
  • Location: San Diego, CA
  • Status: offline
Re: PICKit 4 - marketing vs development philosophy 2018/10/09 06:25:11 (permalink)
0
I do miss the bond out processor modules, but not the expense of buying new ones for each new chip and revision.
I wonder if it might be possible to get near the same features with the addition of a little jtag circuitry inside the chip without adding too much cost to the chip ?
#3
Jim Nickerson
User 452
  • Total Posts : 5280
  • Reward points : 0
  • Joined: 2003/11/07 12:35:10
  • Location: San Diego, CA
  • Status: offline
Re: PICKit 4 - marketing vs development philosophy 2018/10/09 06:38:38 (permalink)
0
I was thinking about something similar to https://www.xilinx.com/su...ion/chipscope_icon.pdf
#4
NorthGuy
Super Member
  • Total Posts : 5051
  • Reward points : 0
  • Joined: 2014/02/23 14:23:23
  • Location: Northern Canada
  • Status: offline
Re: PICKit 4 - marketing vs development philosophy 2018/10/09 07:59:54 (permalink)
0
JANickerson
I was thinking about something similar to https://www.xilinx.com/su...ion/chipscope_icon.pdf



Xilinx's ILA is just a design running on regular FPGA fabric and connected to the fabric under test.
 
The PIC debugger is somewhat similar. When the code halts, the debug executive starts, and then it can transfer information (anything you wish) to the debugger.
 
Of course, PIC16 is not FPGA and it cannot execute debug executive in parallel to your main code. Nor can it do anything at FPGA speeds. Anyway, Logic Analyzer requires much faster speeds than the PIC can offer.
 
PIC24E/C and older PIC32MX have separate debugging hardware which you can talk to, so you can read/write variables (but not SFRs) without halting the CPU. This is a big step forward, but still far from enough for Logic Analyzer.
 
JTAG boundary scan available on many PICs is the closest you can get to getting real-time pin data, but it's still way slower than you would need for a Logic Analyzer. And often JTAG interferes with debugger, so you may not be able to run both anyway.
 
#5
CDS
New Member
  • Total Posts : 12
  • Reward points : 0
  • Joined: 2018/09/27 22:20:44
  • Location: 0
  • Status: offline
Re: PICKit 4 - marketing vs development philosophy 2018/10/09 16:18:07 (permalink)
0
I dont really blame PICKit for its limited debug features since it is a nice gadget at a budget price. Actually, i perefer it over the Real-ICE that i own too and which in some ways is a real pain in the ass. However, debugging with only one breakpiont is a bit too poor sometimes.
Once you hold at a breakpoint you usually have to single step for the next few lines, right? Unfortunately, you cannot even step over unless you release your one and only breakpoint. Real-ICE and ICD can do better in this regard, isnt it?
So i guessed, if we have a 300 MHz core in there, we could have simulated a little bit of their features in software couldnt we?
Instead of wasting those resources for an Atmel support which literally nobody needs.
 
Honestly, are we going to try our first steps on AVR territory next time just because we bought a 40$ PICKit 4 device?
I rather dont think so.
post edited by CDS - 2018/10/09 16:31:43
#6
qhb
Superb Member
  • Total Posts : 7158
  • Reward points : 0
  • Joined: 2016/06/05 14:55:32
  • Location: One step ahead...
  • Status: online
Re: PICKit 4 - marketing vs development philosophy 2018/10/09 16:24:38 (permalink)
0
CDS
...
Once you hold at a breakpoint you usually have to single step for the next few lines, right? Unfortunately, you cannot even step over unless you release your one and only breakpoint. Real-ICE and ICD can do better in that regard, isnt it?

Yes and no.
The number of HARDWARE breakpoints in a device is controlled by the PIC, not the debugger.
If your device supports SOFTWARE breakpoints, then switching to them will give you unlimited breakpoints, but it will wear your FLASH memory a bit quicker.
 

Worst forum problems are now fixed, but the damn firewall is still there.
#7
CDS
New Member
  • Total Posts : 12
  • Reward points : 0
  • Joined: 2018/09/27 22:20:44
  • Location: 0
  • Status: offline
Re: PICKit 4 - marketing vs development philosophy 2018/10/09 17:45:58 (permalink)
0
@qhb: Sure, im aware of this hardware- software-breakpoint thing.
But when we develop for big quantities we have to choose a device that fits the job regardless if it offers some nice-to-have features or not. In the past years, i did a ton of projects on 16F882/883 which are especially annoying because they only break at the next instruction which often is in the next subroutine.
Only recently i switched to 16F18854 as a new favorite because of its awesome core independant peripherals.
However, even 16F18854 is a pretty new device, it features once again only one hardware breakpoint.
 
If it wasnt for the money, i would like to switch again to the brandnew 16F18455 family simply because of its
3 hardware-breakpoints and its 12-bit AD. Unfortunately, it starts with a 14k flash model which is a total waste for the most small and medium complex projects.
 
i would really love to see a 16F18454 as a smaller, 7k flash member of that family.
 
post edited by CDS - 2018/10/09 18:04:29
#8
NKurzman
A Guy on the Net
  • Total Posts : 16442
  • Reward points : 0
  • Joined: 2008/01/16 19:33:48
  • Location: 0
  • Status: offline
Re: PICKit 4 - marketing vs development philosophy 2018/10/09 18:37:18 (permalink) ☼ Best Answerby CDS 2018/10/10 11:43:09
0
Check the Family Tree.  Sometimes the Bigger chips in the Family have more Break points.  You can put those in the a Few Boards to Debug with.
#9
CDS
New Member
  • Total Posts : 12
  • Reward points : 0
  • Joined: 2018/09/27 22:20:44
  • Location: 0
  • Status: offline
Re: PICKit 4 - marketing vs development philosophy 2018/10/10 11:45:18 (permalink)
0
That's indeed a great hint. I'll keep that in mind and will have an eye on it.
Thanks alot..
#10
Jump to:
© 2018 APG vNext Commercial Version 4.5