Who's afraid of the PIC32MX695F512H ?

[blueroomelectronics]

I'm putting together an Ethernet based HA controller and it was going to use a much less frightening PIC18F87J60 microcontroller.

The PIC32MX695F512H (MIPs 32bit core) has boatloads more Flash, RAM, I/O than any 8 bit and is available in an 64pin TQFP plus it's also cheap at $12 qty one.
Some highlights
512k Flash, 128k RAM
Ethernet
OTG USB (can act as a host or slave)
Six UARTs
HW RTCC (handy)

Have you ever wanted to venture into 32 bit PIC territory?
**broken link removed**

 

[EN0]

I'm afraid! At least, for now, that is. I expect once I become more familiar with PIC architecture and programming in general, my fears will dampen.

 

[psecody]

I have wanted to try them for a while but as long as I'm a broke college student I'm probably not going to venture that way, plus I don't even feel competent enough with the smaller pics to do anything much other than display stuff on LCD's. One day though, one day.

 

[blueroomelectronics]

It's a monster compared to its 8 bit cousins. The hard part would most likely be the Ethernet & USB software, else it would appear no more difficult than any other microcontroller.

It's also pretty cheap, about the same price as most 8 bit microcontrollers.

 

[EN0]

I'm surprised that the price is so close to the 8-Bit µC's.

Out of curiosity, how much have you done with the PIC32MX695F512H or similar architecture? What difficulties did you encounter?

 

[BeeBop]

Love the thread title Bill!
I've got the Microchip intro board. and am having fun with it, when work permits me.

I think you are right; no more difficult than any other micro, and yepper, tons more of what makes a microcontroller, a microcontroller!

 

[blueroomelectronics]

Zero with the MIPs core PICs. But plenty with the 8 & 16 bit PICs. Seems they get easier to program as you move up the line. I squarely blame the popularity of AVR micros are because the old 16F series were not the friendliest to get started with and not very compiler friendly because of the limited stack.

 

[DirtyLude]

$12 is about what it should cost. The LPC17xx in 512k flash is equivalent. A few more peripherals, less RAM... fairly similar.

 

[AtomSoft]

You should have noted: 80MHz, 1.56 DMIPS/MHz

Does that mean its : 124.8 DMIPS? which is nice especially for a PIC. Do you still have to use linker files and stuff for these? Like is there a boundary when making arrays and buffers? Like can i have a 1000 Byte buffer?

I switched to ARM7 because of things like that. I can simply create a 1152 Byte buffer like for my LCD and fill it up via Pointer or direct like MyVar[100] = x;

I never seen code for a PIC32 but would love to get more info on it. $12 is a good price for what you get too!

 

[be80be]

There not harder to code you just have to read the datasheet more Here my baby **broken link removed**

The LPC1769/68/67/66/65/64/63 are ARM Cortex-M3 based microcontrollers for
embedded applications featuring a high level of integration and low power consumption.
The ARM Cortex-M3 is a next generation core that offers system enhancements such as
enhanced debug features and a higher level of support block integration.
The LPC1768/67/66/65/64/63 operate at CPU frequencies of up to 100 MHz. The
LPC1769 operates at CPU frequencies of up to 120 MHz. The ARM Cortex-M3 CPU
incorporates a 3-stage pipeline and uses a Harvard architecture with separate local
instruction and data buses as well as a third bus for peripherals. The ARM Cortex-M3
CPU also includes an internal prefetch unit that supports speculative branching.
The peripheral complement of the LPC1769/68/67/66/65/64/63 includes up to 512 kB of
flash memory, up to 64 kB of data memory, Ethernet MAC, USB Device/Host/OTG
interface, 8-channel general purpose DMA controller, 4 UARTs, 2 CAN channels, 2 SSP
controllers, SPI interface, 3 I2C-bus interfaces, 2-input plus 2-output I2S-bus interface,
8-channel 12-bit ADC, 10-bit DAC, motor control PWM, Quadrature Encoder interface,
four general purpose timers, 6-output general purpose PWM, ultra-low power Real-Time
Clock (RTC) with separate battery supply, and up to 70 general purpose I/O pins.
The LPC1769/68/67/66/65/64/63 are pin-compatible to the 100-pin LPC236x
ARM7-based microcontroller series.

 

[Mr RB]

I read the datasheet and wasn't impressed. It looked overly complex to set up and doesn't seem to be that well supported with code and open-source projects people have made. Compilers are less common too.

A real step up from say a PIC24 would be something like a Xmos or Propeller with lots of parallel computing power and tons of support code and projects (more so for the Propeller but Xmos is catching up fast).

I love Microchip products but to me the PIC32 has missed the boat. It's too expensive, fussy and complex for smaller apps, and if you want a real powerful micro (massive MIPS etc) there are much better options out there.

For Microchip to "catch the boat" again I think they need to forget about adding 200 complex peripherals, and instead make something with much less to set up, and fast (800 MIPS?) and craploads of RAM and ROM. With that much power you can do the peripherals in code, which is so much more versatile and powerful.

 

[blueroomelectronics]

How is it too expensive? Digikey sells it for $12 CDN qty one. Massive MIPs what for? I'd prefer offloading UARTs and timing sensitive peripherals to hardware rather than bloaty code that needs constant attention. For a home automatic or HVAC controller you don't need that much horsepower but you do want low power.
It also can be connected with little more than a power supply and a handful of decoupling caps as everything is on the chip.

 

[3v0]

Up to now I have been just watching here.

As for price Newark in the US has it for $8.44. MicrochipDirect.com may have them at about the same price.

I have not looked into the compiler issue.

My thinking is along the same line as Bills.

 

[be80be]

Hay you can even code in asm and from what I seen you could code on a linux box for free in C plus You can use C32 it's as easy a C18 I think.

I would like to see some one do the same as Mbed You just write code with it and it works. That would make C like 3v0 thinks is. Chips like this need to be less hardware code writing and more getting down to businesses

You should be able to just write ADC_read(pinA0) with out the bull of setting registers

 

[blueroomelectronics]

It looks like all you'd need is a PICkit3 or ICD3 to program the beast. There is a free academic C compiler as usual too. So the cost of getting started is quite reasonable (a PICkit3 is about $50).

MIPs and ARM cores have been locking antlers for ages.

To move beyond that I'd consider an embedded Linux solution.

 

[RMMM]

The only 32pit PIC I have used is UBW32.

Even without the bitloader, it is fairly easy to use and not too much of a leap from 8bit PICS.

Using the same environment, it is not too complicated.

Get one and act as if it is the ONLY uC available. Grab a template, and the datasheet and have at it.

I do agree with Mr RB. I would like to see a more "software" dependent powerhouse.
Replace the hardware peripherals with a bunch of memory and let the designer DESIGN the peripherals in software that they need.

Crank up on the MIPS and let us have at it. Releasing software versions of the peripherals with their "pay-for" compilers and that would be a hugely powerful and configurable uC.

Click on the boxes to add a software peripheral to the linker, and the functions could act just as you were using hardware. blueroomelectronics is correct there where it could lead to bloated code, but I think having the power to get through the bloat and function on-time is the important part.

 

[nsaspook]

I read the datasheet and wasn't impressed. It looked overly complex to set up and doesn't seem to be that well supported with code and open-source projects people have made. Compilers are less common too.

A real step up from say a PIC24 would be something like a Xmos or Propeller with lots of parallel computing power and tons of support code and projects (more so for the Propeller but Xmos is catching up fast).

I love Microchip products but to me the PIC32 has missed the boat. It's too expensive, fussy and complex for smaller apps, and if you want a real powerful micro (massive MIPS etc) there are much better options out there.

For Microchip to "catch the boat" again I think they need to forget about adding 200 complex peripherals, and instead make something with much less to set up, and fast (800 MIPS?) and craploads of RAM and ROM. With that much power you can do the peripherals in code, which is so much more versatile and powerful.

32 bit controllers, what can you say. The others guys have one, so you have to have one. It gets you in the door to sell them something else.

 

[3v0]

I favor distributed control. Where is this going with talk of yet more speed and memory. How many wires/sensors and how much sole authority do you really want to put in one uC ?

 

[Mr RB]

Blueroomelectronics-

Massive MIPs what for? I'd prefer offloading UARTs and timing sensitive peripherals to hardware rather than bloaty code that needs constant attention. For a home automatic or HVAC controller you don't need that much horsepower but you do want low power.

Well that is my point. If you don't NEED huge horsepower than a PIC18 or PIC24 is plenty! I have an entire 3-axis CNC machine running from a $1.50 18F2221 doing 3-axis simultaneous accel/decel in very fine resolution cartesian coordinates, while running all the comms too. Something like home automation tasks should be well in the realm of the latest big pinout 18F.

And if you actually start to actually need the amount of horsepower of a PIC32 then there some other very impressive options, that don't force you to buy 6 USARTs etc. If you code peripherals in software you pay for less chip and you can code exactly what you need that do things they way you want.

Part of what I object to with the PIC32 is being forced to pay for all the onboard peripherals AND pay the costs of setting them all up, and the limits of peripherals designed by someone else to suit "everyone". When all I wanted was big mips and RAM. Its like being forced to buy the full pitcrew trailer as well, when all you wanted was the Ferrari. Other microcontroller manufacturers are taking a smarter approach, or starting to.

 

[blueroomelectronics]

Actually it's the multiple UARTs that are the most appealing for the application I have in mind. Built in Ethernet and OTG USB for less than an IMO overpriced FT232R.
Toss in 512k Flash, 256k RAM and an RTCC seems pretty sweet for the price of a Big Mac combo.
I don't need the horsepower and would be delighted if they crammed all those features into an 18F part.
**broken link removed**