HELP! What MCU (greater than 8-bits of smarts) is easiest to graduate to?

[22swg]

16 better than 8 ? Suzy.
http://digitaldiy.io/forum/general-electronics/2396-a-diy-pic24h-dev-board#16966

[miguelvp]

Another thing about the post above using 4 PWMs at once. That was a 4100 so no UDBs, here is the report when programming a 4200 (CY8C4245PVI-482), same capabilities as the one in the prototype board, other than less pins. (24 vs 36)

Code: [Select]

Resource Type                 : Used : Free :  Max :  % Used
============================================================
Digital clock dividers        :    0 :    4 :    4 :   0.00%
Pins                          :    6 :   18 :   24 :  25.00%
UDB Macrocells                :    0 :   32 :   32 :   0.00%
UDB Unique Pterms             :    0 :   64 :   64 :   0.00%
UDB Datapath Cells            :    0 :    4 :    4 :   0.00%
UDB Status Cells              :    0 :    4 :    4 :   0.00%
UDB Control Cells             :    0 :    4 :    4 :   0.00%
Interrupts                    :    4 :   28 :   32 :  12.50%
Comparator/Opamp Fixed Blocks :    0 :    1 :    1 :   0.00%
SAR Fixed Blocks              :    0 :    1 :    1 :   0.00%
CSD Fixed Blocks              :    0 :    1 :    1 :   0.00%
CapSense Blocks               :    0 :    1 :    1 :   0.00%
8-bit CapSense IDACs          :    0 :    1 :    1 :   0.00%
7-bit CapSense IDACs          :    0 :    1 :    1 :   0.00%
Temperature Sensor            :    0 :    1 :    1 :   0.00%
Low Power Comparator          :    0 :    2 :    2 :   0.00%
TCPWM Blocks                  :    4 :    0 :    4 : 100.00%
Serial Communication Blocks   :    0 :    2 :    2 :   0.00%
Segment LCD Blocks            :    0 :    1 :    1 :   0.00%

Not a single UDB was used.

Here is the schematic of the modules used

[miguelvp]

But wait! There is more!

6 PWMs using 2 UDBs plus the 4 TCPWMs and still 2 UDBs left.

Code: [Select]

Resource Type                 : Used : Free :  Max :  % Used
============================================================
Digital clock dividers        :    1 :    3 :    4 :  25.00%
Pins                          :    8 :   16 :   24 :  33.33%
UDB Macrocells                :    9 :   23 :   32 :  28.13%
UDB Unique Pterms             :    8 :   56 :   64 :  12.50%
UDB Total Pterms              :    8 :      :      :
UDB Datapath Cells            :    2 :    2 :    4 :  50.00%
UDB Status Cells              :    1 :    3 :    4 :  25.00%
            StatusI Registers :    1
UDB Control Cells             :    1 :    3 :    4 :  25.00%
            Control Registers :    1
Interrupts                    :    5 :   27 :   32 :  15.63%
Comparator/Opamp Fixed Blocks :    0 :    1 :    1 :   0.00%
SAR Fixed Blocks              :    0 :    1 :    1 :   0.00%
CSD Fixed Blocks              :    0 :    1 :    1 :   0.00%
CapSense Blocks               :    0 :    1 :    1 :   0.00%
8-bit CapSense IDACs          :    0 :    1 :    1 :   0.00%
7-bit CapSense IDACs          :    0 :    1 :    1 :   0.00%
Temperature Sensor            :    0 :    1 :    1 :   0.00%
Low Power Comparator          :    0 :    2 :    2 :   0.00%
TCPWM Blocks                  :    4 :    0 :    4 : 100.00%
Serial Communication Blocks   :    0 :    2 :    2 :   0.00%
Segment LCD Blocks            :    0 :    1 :    1 :   0.00%

Although the last two share the same interrupt.


Maybe I can squeeze another 2, although the datasheet says these PWMs only use one UDB maybe they share the datapath cells which this last one used two of them(one per output pin I guess). Still I should be able to do 8 total.



And yeah, the 2 new ones are still 16 bits:


But I did move to the 4200, the 4100 doesn't have UDBs so it can only have 4.

[miguelvp]

Well 8 count for 16 bits PWMs on the PSoC 4200, but now I'm out of datapaths on the UDB and maxed out on TCPWMs blocks as well:

Code: [Select]

Resource Type                 : Used : Free :  Max :  % Used
============================================================
Digital clock dividers        :    1 :    3 :    4 :  25.00%
Pins                          :   10 :   14 :   24 :  41.67%
UDB Macrocells                :   16 :   16 :   32 :  50.00%
UDB Unique Pterms             :   15 :   49 :   64 :  23.44%
UDB Total Pterms              :   15 :      :      :
UDB Datapath Cells            :    4 :    0 :    4 : 100.00%
UDB Status Cells              :    2 :    2 :    4 :  50.00%
            StatusI Registers :    2
UDB Control Cells             :    2 :    2 :    4 :  50.00%
            Control Registers :    2
Interrupts                    :    6 :   26 :   32 :  18.75%
Comparator/Opamp Fixed Blocks :    0 :    1 :    1 :   0.00%
SAR Fixed Blocks              :    0 :    1 :    1 :   0.00%
CSD Fixed Blocks              :    0 :    1 :    1 :   0.00%
CapSense Blocks               :    0 :    1 :    1 :   0.00%
8-bit CapSense IDACs          :    0 :    1 :    1 :   0.00%
7-bit CapSense IDACs          :    0 :    1 :    1 :   0.00%
Temperature Sensor            :    0 :    1 :    1 :   0.00%
Low Power Comparator          :    0 :    2 :    2 :   0.00%
TCPWM Blocks                  :    4 :    0 :    4 : 100.00%
Serial Communication Blocks   :    0 :    2 :    2 :   0.00%
Segment LCD Blocks            :    0 :    1 :    1 :   0.00%

But still have UDB resources available other than not external. This is the schematic the last 4 share 2 interrupts, I can make the last four single output and each have their own interrupt since there are plenty of interrupts left.

Also I could have 7 PWMs and an  TimerCounter, or 6PWMs timer counter and quadrature decoder.
On top of that I could let the first 4 TCPWMs unconfigured and configure them on the go programmatically.

Anyways, this is the schematic:


Here they are all off (or barely on):


And here they are brighter:

[miguelvp]

SuzyC,

You might be interested in this project with full source:

Flying PSoC4:
http://www.cypress.com/?app=forum&id=2492&rID=93005

It's a proof of concept that runs 4 servos directly from 4 pots. And will work on the Pioneer or the KIT-049-42xx prototype board like the ones you have.

I haven't tried it yet, but it does compile. It also has a multi-page schematic so when you open the top level design, make sure you look at all the pages.

Edit, it leaves plenty of space for other things:
Flash used: 5350 of 32768 bytes (16.3 %).
SRAM used: 1564 of 4096 bytes (38.2 %).

And plenty of resources left including 2 full UDB datapath cells and 11 macrocells.

Code: [Select]

Resource Type                 : Used : Free :  Max :  % Used
============================================================
Digital clock dividers        :    1 :    3 :    4 :  25.00%
Pins                          :   13 :   23 :   36 :  36.11%
UDB Macrocells                :   21 :   11 :   32 :  65.63%
UDB Unique Pterms             :   24 :   40 :   64 :  37.50%
UDB Total Pterms              :   24 :      :      :
UDB Datapath Cells            :    2 :    2 :    4 :  50.00%
UDB Status Cells              :    1 :    3 :    4 :  25.00%
            StatusI Registers :    1
UDB Control Cells             :    1 :    3 :    4 :  25.00%
            Control Registers :    1
Interrupts                    :    2 :   30 :   32 :   6.25%
Comparator/Opamp Fixed Blocks :    0 :    2 :    2 :   0.00%
SAR Fixed Blocks              :    1 :    0 :    1 : 100.00%
CSD Fixed Blocks              :    0 :    1 :    1 :   0.00%
CapSense Blocks               :    0 :    1 :    1 :   0.00%
8-bit CapSense IDACs          :    0 :    1 :    1 :   0.00%
7-bit CapSense IDACs          :    0 :    1 :    1 :   0.00%
Temperature Sensor            :    0 :    1 :    1 :   0.00%
Low Power Comparator          :    0 :    2 :    2 :   0.00%
TCPWM Blocks                  :    4 :    0 :    4 : 100.00%
Serial Communication Blocks   :    0 :    2 :    2 :   0.00%
Segment LCD Blocks            :    0 :    1 :    1 :   0.00%


[Kjelt]

miguel it looks like you better can start your own topic, people interested in the psoc4 shall not find it here. And it looks like Suzy already has left the psoc4 boat.

[paulie]

I thought so too but these naughty topics don't always behave as well as we sometimes plan. Looks like "Easiest to Graduate" is the new "PSOC Examples". Miguel is very enthusiastic about his true love. I didn't mind at all to see these posts show up in the $1 thread because I learned a lot and it opened my eyes to another potential member of the club.

Speaking of which I've done some experiments on the 1.5k ADC subject. Noticing that with 100k pot pin 1.7 was giving accurate results when 1.6 was way off I looked into it. Main difference was a 1uf cap so I tried putting one on 1.6. Sure enough reading improved in a big way. Looks like the impedance is mainly reactive (probably high series R before S/H too) and tacking a few uf on can save using those unbelievably hard to come by op amp blocks. Of course this won't do for high sample rate DSP type apps but that's not likely a problem reading robot position sensors.

ps.Here's a photo of my ADC test setup:

[miguelvp]

miguel it looks like you better can start your own topic, people interested in the psoc4 shall not find it here. And it looks like Suzy already has left the psoc4 boat.

That might be the case, but the topic was about controlling multiple motors and finding another MCU capable of doing so.

So SuzyC gave up too early on the PSoC, when it's more than capable, and I should just say nothing?

Nah, just because of a few road bumps, one should not give up. Actually if the OP didn't need them all to be 16 bits, the PSoC 4 prototype board can do 4 16bit PWM and 12 8 bit ones, for a total of 16 motor controls in a single $4 prototype board.

So I won't post the 16 PWMs in a single board example, but just because one can't see what a chip can do it doesn't mean they can claim is useless and won't fit that specific need, I'm just showing it does with plenty of spare capabilities.

But I already said my peace on how capable that chip is, but it's my prerogative to prove that it is. After all it's not my subject so I really shouldn't need to do my own topic because I have no need to run multiple motors, at least not yet

Edit: I haven't even gone to how you can take one of those components and modify them to your own needs. Meaning I can make my own custom PWM based on the ones supplied with full source and resources, just like if I started the component from scratch myself. I do wish it was an M3 with lower clock cycles or that the 5LP was more affordable, but the 4 with the M0 is more than capable.

[SuzyC]

Suzyc does not give up, Suzyc fights to the last.
Suzyc sometimes gets discouraged by frustration.

Any device that fails to clearly document a product and forces the a customer to review  poorly developed documentation that ultimately force even a serious programmer to resort to an endless game of guessing and googling is a POS.

Life is too short for this.


Ok, once more why the PSOC4 is a POS:

I suddenly see PWM's expanding from my own limit of two two-output 16-bits PWM UDBs to a single non-UDB block of maybe up to eight PWM's. When I add the A2D block and a counter UDB, I am limited to one 2-output 16-bit PWM and somehow it allows me to add one more 8-bit  PWM and then if I change it to a 16-bit PWM I get the "need 5  you only got 4 " gotcha!

When I search the resources of my chip I do not find these hybrid multi-more PWM enabled PWM blocks.  Why is it that the Great Creator and Cypress gives me no hint they even exist or can be created  And while there are those that undoubtedly created these blocks so essential to my MMMMM's, I have not been guided by Cypress PWM UDB documentation on how to implement them without exhausting my UDB count .  Cypress's chain of information seems to fail me in that it so well succeeds in hiding the info I need. With Microchip, I can get this info just by carefully studying the chip Spec Sheet.

So, that's why the PSOC4 is a POS.

[Laurynas]

Those multiheaded PWMs are user libraries and while it would be nice for Cypress to add them to Creator, I think one should not fault them that they don't have all user created content in their software.
Myself i used those components in a 20 digital input + 12 PWM output board. It would have been faster to have them in Creator, but 5mins of googling did the job.

[miguelvp]

Actually they are both installed by default, I didn't use any user created ones. One is called PWM the other TCPWM and they are both in the same toolbox.

The only one that is not in there and you can get from the user community is the 3 output 8 bit one. Did you even looked at the quadcopter link and code in there?

but it's obvious I'm wasting my time, so Uncle.

Edit: This is how obvious the two components are, right in plain view. Top one takes no UDBs the bottom one takes 2 UDB paths (for 16 bit transfers) per. Hook Interrupts to them, clear the interrupt when called and set whatever needs to be set in the interrupt, it really can't get any simpler than that.

[dannyf]

even a serious programmer to resort to an endless game of guessing and googling is a POS.

I am not a serious programmer so the above statement definitely doesn't apply to me.

I just wanted to point out that no one is an expert on anything day 1. You have to learn, serious programmer or not. When you try to learn, you use every resource available to learn, googling included. I am not quite understanding your googling aversion.

From afar, the situation looks like your frustration with learning, more so than with the chip. Frustration is unfortunately part of the learning. Either you tough it up and come out on the other end as a winner, or you get back to things you are comfortable with.

I don't know that chip that well - I actually find it quite limiting and its programming approach unappealing to me. But I find the concept of mcu + cpld quite useful in some applications. So it seems to me a lot more productive if you focus on getting the chip to work for your application than complaining about its limitations.

[miguelvp]

I do understand frustration, but even after 30 years of programming, when I try something new, I feel no problem asking a coworker for help if he/she has already researched that topic in a system and other senior programmers come to me to ask me questions on things I already researched.

For example XboxOne is a total different beast, it's hard sometime to find the right documentation, if any is available, etc, but we still push along regardless and you find senior and junior programmers alike helping each other. Maybe I'm more adaptable to changes since in my field things change constantly and maybe because of that I find it easy to navigate through the documentation provided by Cypress, which is maybe overwhelming but surely not lacking.

So If you need clarification, you can always ask. It's easier to ask how do I fit four 16 bits PWMs an ADC and a counter than just saying, this is a piece of shit and I know it can't be done so I'm not going to even bother asking, because if I can't figure it out then is useless and I'll complain about it instead of seeking for help.

So if you just need  four 16 bits PWMs an ADC and a Counter then use 2 single channel hardware based TCPWMs, 1 dual channel 16 bits UDB based PWM (it uses 2 UDB datapaths 8 bits each out of 4 available) Then use a TCPWM based timer counter that uses no UDBs and you can fit your ADC with ease.

Of course clearing the interrupt is different in the TCPWM than in the PWM.

[paulie]

Did you even looked at the quadcopter link and code in there?

Quadcopter? Did somebody say quadcopter? Quadcopter what? Is there a link to quadcopter? Is there a link to the quadcopter link?

After playing with over a dozen different ARM part numbers last couple weeks I'd have to say PSOC4 was not among the easiest. Some took minutes, some took hours. PSOC was not one of the "minutes" ones. PSOC would have been a lot more attractive to me if it weren't for the demo board fake programmer and lack of basic easy to use flash utility. Still it was an interesting experience and well worth the $4 demo board if not the $10 spent on chips.

[miguelvp]

Did you even looked at the quadcopter link and code in there?

Quadcopter? Did somebody say quadcopter? Quadcopter what? Is there a link to quadcopter? Is there a link to the quadcopter link?

After playing with over a dozen different ARM part numbers last couple weeks I'd have to say PSOC4 was not among the easiest. Some took minutes, some took hours. PSOC was not one of the "minutes" ones. PSOC would have been a lot more attractive to me if it weren't for the demo board fake programmer and lack of basic easy to use flash utility. Still it was an interesting experience and well worth the $4 demo board if not the $10 spent on chips.

Don't get too excited about about it. It's just a proof of concept to drive 4 servos based on pot inputs.

The link is in this post:
https://www.eevblog.com/forum/microcontrollers/help!-what-mcu-(greater-than-8-bits-of-smarts)-is-easiest-to-graduate-to/msg529189/#msg529189

And following the link here is what the guy that did the project says about what the project contains:

I was afraid that in a recent post I bit off more than I could chew when I promised HL a "PSoC 4 controlling a quad copter using Calman filters". So I started to build a proof-of-concept which is not ready yet but it turns out that it will really fit. A bit tight, using up most resources, but will fit. Not even Calman, but there is something even better.
 
The copter needs controlling 4 motors using a standard rc-interface (1ms to 2ms PWM) and as input a MPX sum-signal. RC-hobbyists know that interfaces.
When building a radio-controlled-device you will need something to easily control your RC-environmentTo get you acquainted to this matter (and my style of solving problems like that) I bult a device to control directly 4 servos with 4 potentiometers.
Project runs on Pioneer board or on KIT-049-42xx,
 
Bob

As for programmer, it does support a simple programmer and Keil uVision supports the chips as well. Not sure about the UDB stuff, but you can export a Creator project to keil and it should build and debug with the standard uVision tools as long as you use SWD.

[paulie]

OK, so more like "hint of concept". I suspected trouble when he consistently miss-speled Kalman. At least he didn't claim "super easy to program" like dannyf. It would be a miracle if anybody managed to single handedly develop multicopter code on a minimum chip like this that worked.

It would also be impressive if somebody put together an easier way to turn the $4 board into a programmer or come up with a flash utility comparable to ST flash demo, STVP or NXP Flashmagic.

[miguelvp]

It would also be impressive if somebody put together an easier way to turn the $4 board into a programmer or come up with a flash utility comparable to ST flash demo, STVP or NXP Flashmagic.

Done: https://www.eevblog.com/forum/microcontrollers/psoc-examples/msg522554/#msg522554

Edit: Not my work other than the tutorial on the steps to follow.

[paulie]

Wow... don't know how I missed that. Spending too much time in the ghetto I guess. With those photos and details it looks a lot better than the original Cypress link you put in the old thread.

My problem is not having a Pioneer or Miniprog3 and can't justify spending $50 or $100 for casual poking around. I tried using the STlink dongles but that was a total failure.  If we knew of somebody who did manage it might be worth another try.  Or maybe you could do a big favor and flash it for me if I sent my $4 board with return postage? I have several of the bare chips and blank breakout boards like yours and would love to make use of them.

[miguelvp]

You don't need a miniprog nor a pioneer to convert the $4 prototype board into a programmer.
It's a bootloadable project.

Edit: There are 3 posts

first one is the program for the target board, and how to wire things up.

2nd one is how to setup the $4 board as a programmer
https://www.eevblog.com/forum/microcontrollers/psoc-examples/msg522555/#msg522555
3rd one is how to use the programmer to program the target board
https://www.eevblog.com/forum/microcontrollers/psoc-examples/msg522556/#msg522556

[paulie]

Are you serious... how dense can I be? So I load the programmer firmware into the $4 board and then use the python script to flash ANY of the bare chips? Simple as that? The $4 board never loses it's boot ROM?

[miguelvp]

Are you serious... how dense can I be? So I load the programmer firmware into the $4 board and then use the python script to flash ANY of the bare chips? Simple as that? The $4 board never loses it's boot ROM?

You got it. and you can reflash it back to another bootloadable image if you need to use it for something other than a programmer.

[neslekkim]

but it's obvious I'm wasting my time, so Uncle.

No, this is interresting reading, until I have time to connect my psoc-sticks..

[paulie]

You got it. and you can reflash it back to another bootloadable image if you need to use it for something other than a programmer.

Great... At first the mention of "Stamp" had me thinking a Parallax board was required so i went no further. Then all that talk about Pioneer/Miniprog3 and "bricking" the $4 board stalled me. Now I have a better picture and will report back on progress. Thanks for clearing things up for the "special" class guy.

[true]

Any device that fails to clearly document a product and forces the a customer to review  poorly developed documentation that ultimately force even a serious programmer to resort to an endless game of guessing and googling is a POS.

Yeah, I might agree with this statement. But for what you later write about, it's not the fault of the documentation. I'm sorry the lack of reading ability is hindering your ability to migrate to a more powerful platform. I doubt parts from Cypress, NXP, ST, TI, SiLabs or any other manufacturer will make you happy. You are probably best off with your PIC16s.

I agree with dannyf's reply and miguelvp's subsequent reply to you about this.

When I search the resources of my chip I do not find these hybrid multi-more PWM enabled PWM blocks.  Why is it that the Great Creator and Cypress gives me no hint they even exist or can be created

It's too late I'm sure, but it says what you need right in the datasheet, as plain as it gets, giving the name of the resource right on page 1. A description of the resource is on page 7. It isn't buried.

As for the resources on the chip, what you described is again an example of the lack of understanding the resources of the device, which can be read on page 1 of the datasheet. I understood it in minutes. miguelvp likely understood it in minutes. Developers using this chip in production products (some of which you may have used or even continue to use) likely read the datasheets to see if the hardware would work, and therefore understood it in similar time. We all understood what peripherals PSoC4 has because this information is very clear in the datasheet. And some of us even knew it had programmable digital logic that could help us in other ways. But if you didn't know that, it says that in the datasheet too.

If you aren't understanding how to use flexibility of the non-CPU parts of "greater than 8-bits of smarts" chips to do more than your PIC16 micros, I don't know how to help you. PSoC isn't for you. STM32, LPC, TM4C, MSP430, SAM, EFM32, and hell maybe even PIC32 probably aren't for you either, mostly because their datasheets don't look like and read like Microchip PIC16 datasheets. If you insist on not using PIC16, maybe look at PIC24, or possibly even dsPIC?

With Microchip, I can get this info just by carefully studying the chip Spec Sheet.

Same goes for Cypress. Stop looking for UDB. Look for what you want, which is PWM. The name of the resource is on page 1. The description of the resource is on page 7.

So, that's why the PSOC4 is a POS.

I'm sorry your lack of reading comprehension makes you feel that way; with that outlook, you will miss out on a lot of really neat things in life. But you'll probably feel more comfortable hacking with a PIC than any other micro, because other micros you may consider aren't 8-bit PICs.

As I wrote here before, I believe that ultimately the question you originally brought up in this topic is not the correct question. This is because any valid answer cannot be suitable for your purpose, which is to have a PIC16.

I don't know that chip that well - I actually find it quite limiting and its programming approach unappealing to me.

Yeah, I did too, when I also didn't know it well. But then I thought about how I could leverage it and tried it for one project. It really is quite the neat chip in some situations, particularly the 4200 (not sure I see much use to the 4100). Thanks to the UDBs and built in op amps I am saving some money and board space, as well as doing things other micros in this size profile and over twice the price still can't do. Working with and learning about the product, it really seems like a helpful tool for the toolbox. And the $0.29 PSoC4000 will probably see use in future i2c slave projects where I have used more expensive chips that didn't use many hardware resources (though I _do_ wish it had one more TCPWM, but I'm getting off track a bit...) Plus, Cypress is a great company to work with.

I still use STM32, PIC32, or even attiny most of the time but for some projects I can complete them much faster, or have much faster hardware, and sometimes do so at cheaper cost.

[miguelvp]

Well dannyf might be happy about this:

Using Keil uVision as a full programmer debugger with PSoC chips:
https://www.eevblog.com/forum/microcontrollers/psoc-examples/msg529947/#msg529947

And I didn't buy a programmer
No, it's not the $4 protoboard.