One Dollar One Minute ARM Development

[mrflibble]

I don't see the F103 as much superior in every respect, nor do I even see them costing less.
...
F030's for $0.70, yes.

You are right, it's probably not a good idea to compare ST chips in the LPC thread so here we are.

I might be wrong and open to education. Admit not pouring over F030 specs but just listening to ghetto talk. Can you be more specific about the advantages of F030? Also my shopping for that has only been Ebay type sites where they are quite expensive so could use a link to  70 cents.

What now? You say F103 is cheaper than F030. That is easy enough for me to check (and exploit by pressing BUY NOW). Explaining what I like about chip #1 vs chip #2 for a few projects takes far longer.

You have PM with link to $0.70 goodies.

[mrflibble]

Short version of why F030 instead of F103:

For a given cost of the chippies:
number of timers + usable PWM pins + usable ADC channels while still having an USART + debug interface. Basically cost per PWM output + ADC input pair, while still maintaining a decent adc converter per adc channels ratio. PWM has to be center aligned, so that limits the timer to advanced or some of the general timers.

And then all things considered, including but not limited to price, the F030 turns out to be "better" than F103. So in this case better happens to be how many of resource type XYZ can you actually use, while not getting IO pin collisions with that other resource. And I've seen enough MCU's where that turns out to be a depressingly low ratio.

But overall, the F103 datasheet has larger numbers in it. No denying that. It's just that useful pins + used peripherals / $$$ turns out better for F030 in more than zero cases.

Anyways, I like the F103 as well and intend to plonk it in some other projects. As far as I'm concerned it is not so much this is better than that PERIOD. It's just that for one use case one chip turns out better, and for another use case the advantage goes to some other chip. At the moment I mainly muck about on ST chips, and TI to a lesser extent. Only so much time in the day to familiarize yourself with a new vendor's goodies.

On the list of possibly fun to try are those NXP's (have been there for a long time now ) and those damn cute PSOC's. Although to be fair the PSOCs stand a better chance of being ordered since they actually bring some new solutions to the toolbox. LPC would just be more of the same (for me). But that doesn't stop me from reading other people's exploits on those, because who knows what cool stuff they find.

[paulie]

You have PM with link to $0.70 goodies.

Thanks for that. However maybe you overlooked the shipping charges there which bring actual cost to $.90. Still a good deal but I'm spoiled by being lucky to snag 103 for a dime cheaper ($.80). I realize this is not a fair comparison because of quantity and not just a click on "buy now". F0 not being a commodity item is not available through that channel. But maybe you can see why I don't consider those a bargain.

Speaking of comparisons, I want to thank you for motivating me to finally cruise that spec sheet too. Here's a little summary table I spent a couple minutes throwing together:

Code: [Select]

STM32F103C8T6 STM32F030F4P6
M3            M0
72mhz         48MHz
128KB flash   16KB
3 UART        1 UART
36 io         15
20K ram       4K
There's more but probably enough to allow most readers to decide which is the "better" chip. Considering cost is similar and 103 is orders of magnitude more available I've made up my mind.

BTW as far as pin/function mapping I find it hard to imagine the 20 pin crippled chip has advantages over the 48 pin in that area. I spent some time but couldn't find a single example. But we do agree totally on the exception for "more than zero cases". It may be possible the 20 pin has a leg up in terms of PCB routing in some rare design but for the other 99% the 103 would usually do. One can always ignore features not needed but hard to make use of those not present.

ps. LPC parts came in and it was surprising to see the 20 pin SMD package was actually bigger than the 48 pin 103, a  LOT bigger.

[paulie]

This is probably best answered here because someone else just asked a similar question.

Not much more than a dollar each at Mouser? Including shipping cost? Woah!
*ducks for cover*

Yes.

http://www.mouser.com/ProductDetail/NXP/LPC812M101JD20FP/?qs=%2fha2pyFadujd6Yxbwm6s6vWcdu8reqzbi4ydgJDibZaEL%2f0jTMWFtw%3d%3d

In this case I followed the "Fribble Method" of minimizing expense by tacking onto a large order so shipping worked out to about 3 cents each. Even without this total cost would still be a dollar and change. The chips have actually arrived now. Next step lets see if we can manage to program these with same minimum tools as the ST parts.

[paulie]

This is promising. I've hooked one of these up to the PC and it seems to be responding to serial commands as described by SirNick at the beginning of his C/Unix thread. Now to see if the Flashmagic utility sca gave me works to upload a hex file. All I have ATM are ones that were assembled for the ST chips so can't expect it to do anything but it will be a good sign if there are no error messages.

[nctnico]

Hmmm, NXP cared to put a 30MHz ARM microcontroller in a DIP8 package.

[paulie]

PSOCs stand a better chance of being ordered since they actually bring some new solutions to the toolbox.

That's what I thought. I bought the $4 dev board and a few chips then found out there is no bootloader. The break-in-half board is nothing more than a USB/serial so that can't be used to flash them either. More money for a custom programming dongle? Nope. Back in the junk box.

Looks like STM and LPC may be the only solution for low cost ARM experiments.

[mrflibble]

That's what I thought. I bought the $4 dev board and a few chips then found out there is no bootloader. The break-in-half board is nothing more than a USB/serial so that can't be used to flash them either. More money for a custom programming dongle? Nope. Back in the junk box.

Using the ~ $25 pioneer board as programmer has been mentioned a few times by miguelvp. Your call if $25 is acceptable for a dev board that can double as programmer for those $4 boards.

[miguelvp]

PSOCs stand a better chance of being ordered since they actually bring some new solutions to the toolbox.

That's what I thought. I bought the $4 dev board and a few chips then found out there is no bootloader. The break-in-half board is nothing more than a USB/serial so that can't be used to flash them either. More money for a custom programming dongle? Nope. Back in the junk box.

Looks like STM and LPC may be the only solution for low cost ARM experiments.

How much do you want for your junk?

Btw you can use the $4 dev board to program (not debug) the other chips.

http://www.cypress.com/?id=4&rID=98796

Edit: The $25 Pioneer can also debug the code on the chips, but to just program it, the $4 breakout board will do.

Edit2: Also you can use Keil to program (And Debug) the chip:

https://www.eevblog.com/forum/microcontrollers/one-dollar-one-minute-arm-development/msg515630/#msg515630

But if you use ulink make sure you apply power to the chip because it doesn't supply power.

[SirNick]

LOL  Man.. if I ever sell products, I'm going to hire you guys to be my parts shoppers.  I'm too lazy.. I just go to Digikey and pick what I want, and grumble if I'm "forced" to second-source stuff from Mouser. 

Hey, Paulie.. if you or westfw end up with an ASM file for the 812, let me know.  I could use something to test the bootloader code I'm working on, and don't want to get distracted going down the rabbit hole of writing a proper test app just yet.

[paulie]

Btw you can use the $4 dev board to program (not debug) the other chips.

I don't think that's the case. First of you should know the CY8CKIT-049 "programmer" is nothing more than a USB/serial interface and requires the target to have a built-in bootloader. As mentioned last time we discussed this the target chip on the 049 obviously has one but it is not one of the $1 parts included in the promotion.

I spent quite a lot of time investigating this and at best came up with several links explaining how to compile and install a custom bootloader using a programmer that has true SWD flashing capability like the miniprog3 or pioneer. These are quite costly for my purpose. Your link showing how to build a programmer using parallax stamp was similar to others that didn't look cheap or easy to implement either.

Next in conversations with Cypress FAE at digikey was told the 4200 chips I purchased did not have boot code installed. Other more expensive versions do.

Finally cypress documentation AN84858 states:

b. Bootloaders can use any standard communication interface (e.g., USB, I2C, SPI, and UART) to update the firmware. But HSSP uses an SWD or JTAG interface to program the flash. PSoC 4 supports only SWD interfaces.

There were a few more hints in this direction but along with high cost after promotion ($4-$5 each) I decided these were not worth pursuing  as a low cost ARM option for casual experimentation. If you have any convincing evidence to the contrary I would be very appreciative as there's still a few sitting around here gathering dust.

[miguelvp]

Btw you can use the $4 dev board to program (not debug) the other chips.

I don't think that's the case. First of you should know the CY8CKIT-049 "programmer" is nothing more than a USB/serial interface and requires the target to have a built-in bootloader. As mentioned last time we discussed this the target chip on the 049 obviously has one but it is not one of the $1 parts included in the promotion.

Ok, the CY8CKIT-049 is a USB/serial interface and does require the target to have a built-in bootloader, and as you mention the chip built into the 049 has the bootloader programmed.

The link I gave:
http://www.cypress.com/?id=4&rID=98796

Uses that 4200 inside the 049 to program an external chip that doesn't have a bootloader and it's factory virgin, so you can program the 2nd external 4200 (or 4100, or whatever).

The program in there uses the on board 4200 pins P2.0 for SWDIO, P2.1 for SWDCK and P2.4 for XRES_n on the protoboard to connect to the external chip. You hook those to the virgin 4200 on P3.2 for SWDIO, P3.3 for SWCLK and the appropriate XRES pin (depends on what chip you have (pin 33 on the 44-TQFP, pin 30 on the 40-QFN or pin 25 on the 28-SSOP packages).

So I'm not talking about using just the USB-UART part of the board, but the whole prototype board to program the external virgin chip.

I spent quite a lot of time investigating this and at best came up with several links explaining how to compile and install a custom bootloader using a programmer that has true SWD flashing capability like the miniprog3 or pioneer. These are quite costly for my purpose. Your link showing how to build a programmer using parallax stamp was similar to others that didn't look cheap or easy to implement either.

Do you have any current programmers listed here?


If you do then you can use Kiel to program and debug the chip. Otherwise you can use the prototype CY8CKIT-049 42xx board as stated.

Next in conversations with Cypress FAE at digikey was told the 4200 chips I purchased did not have boot code installed. Other more expensive versions do.

Finally cypress documentation AN84858 states:

b. Bootloaders can use any standard communication interface (e.g., USB, I2C, SPI, and UART) to update the firmware. But HSSP uses an SWD or JTAG interface to program the flash. PSoC 4 supports only SWD interfaces.

There were a few more hints in this direction but along with high cost after promotion ($4-$5 each) I decided these were not worth pursuing  as a low cost ARM option for casual experimentation. If you have any convincing evidence to the contrary I would be very appreciative as there's still a few sitting around here gathering dust.

what $1 chips did you get (package information) and what prototype board did you get (41xx or 42xx)?
Because the project in that link is easier on the 42xx board.

The project in that link "4 dollar programmer.zip" should be able to program (not debug) the external virgin $1 chip with any .hex file that is compiled for that external chip.


Edit: for example I did program this chip with the Pioneer, and it was one of those $1 chips, I can reprogram it using the $4 42xx prototype board and document the steps if you like.
It doesn't have a bootloader because I didn't wire a button to activate the bootloader on it nor a bootloader status led for that matter.

[Laurynas]

Btw you can use the $4 dev board to program (not debug) the other chips.

http://www.cypress.com/?id=4&rID=98796

Edit: The $25 Pioneer can also debug the code on the chips, but to just program it, the $4 breakout board will do.

Nice find!
I was using my Pioneer as a programmer, but it's nice to see that usb->serial->1$ PSoC->SWD programmer is implemented.
I wonder if this would work with 4000 series basic chips

[miguelvp]

Btw you can use the $4 dev board to program (not debug) the other chips.

http://www.cypress.com/?id=4&rID=98796

Edit: The $25 Pioneer can also debug the code on the chips, but to just program it, the $4 breakout board will do.

Nice find!
I was using my Pioneer as a programmer, but it's nice to see that usb->serial->1$ PSoC->SWD programmer is implemented.
I wonder if this would work with 4000 series basic chips

As the target yeah, but not as the programmer since you need the UART and the UART SCB mode components which they won't be present in the 4000 series.

I know they are present in the 4200 I hope they are in the 4100 as well.

Edit: BTW, are you using the pioneer as a programmer debugger or just as a programmer? the reason I ask is because there is a way to use it as a miniprog 3 if you are interested.

[miguelvp]

Ok, didn't do the detailed in-depth yet but just wanted to make sure it worked.

Caveat: and a big one! the project is not a bootloadable one so once you program your $4 prototype board you can't restore it unless you have another $4 board or a $25 Pioneer or a $90 miniprog, or other programmers including Altera and Xilinx programmers among the ones supported by Keil, but I only have the Pioneer and the $4 prototype board (2 of them one 42xx the other 41xx and I haven't tried it with the 41xx yet).

So I took the project and built it, boot loaded it into the prototype board.

Then I made a blinky program for the external chip and uploaded the .hex code following the instructions.
wired the VDD, GND, XRES, SWDIO & SWDCLK



And it runs the blink program on the external chip just fine



Again, this will disable the bootloadable code on the prototype kit, I guess I can work on making the $4 programmer code bootloadable so you don't brick the prototype board.

I can recover it because I have the pioneer to override the code, but in any event, once you program the prototype kit as a programmer you can use it indefinitely by changing the python script to point to a different hex file to program your target chip.

Edit2: One more caveat, unless you put the right decoupling caps to use the internal regulator, you have the choice of shorting pins but then you have to feed it 1.8V +- 5% available on the VCCD pin but that is not exposed on the headers and it just goes to C1, I guess you can solder a wire to get the VCCD 1.8 rectified voltage if you decide to short the target chip pins to use the external regulated supply.

But might as well put the decoupling caps so you can feed anything from 1.8V to 5.5V.

[Laurynas]

Nice find!
I was using my Pioneer as a programmer, but it's nice to see that usb->serial->1$ PSoC->SWD programmer is implemented.
I wonder if this would work with 4000 series basic chips

As the target yeah, but not as the programmer since you need the UART and the UART SCB mode components which they won't be present in the 4000 series.

I know they are present in the 4200 I hope they are in the 4100 as well.

Edit: BTW, are you using the pioneer as a programmer debugger or just as a programmer? the reason I ask is because there is a way to use it as a miniprog 3 if you are interested.

4100 series have SCBs so they should work here. But I just wish that someone would write a decent software UART (not just TX that cypress has as standard component.) this way we could make some use of ARM core and get this into 8 pin minimal devices. (not that it has some real cost advantage, just for fun:))
As for pioneer I soldered some wires where the resistors were so it works like a proper debugger.

[miguelvp]

As for pioneer I soldered some wires where the resistors were so it works like a proper debugger.

You'll be interested then in this:
http://www.element14.com/community/thread/27067/l/psoc-5lp-programmer-firmware

On the 2nd page I modified the released 2.03 KitProg code to use the 5LP header to drive the SWDIO and SWDCLK, also not in there I changed the version.h code to force it to 2.08 so that it works with other cypress tools.

http://www.element14.com/community/thread/27067?start=15&tstart=0

[Laurynas]

bookmarked
much more work done than soldering, but looks much better than my solution.

[miguelvp]

bookmarked
much more work done than soldering, but looks much better than my solution.

Not really more work, I just changed two pins on the cydwr file so it will use the 5LP header pins for SWDIO and SWDCLK and the version number, also prevented the bootloader/bootloadable modules to upgrade past version 1.1 and the USBFS to leave them at 2.51 or 2.50 where 2.51 was not an option.

[Laurynas]

bookmarked
much more work done than soldering, but looks much better than my solution.

Not really more work, I just changed two pins on the cydwr file so it will use the 5LP header pins for SWDIO and SWDCLK and the version number, also prevented the bootloader/bootloadable modules to upgrade past version 1.1 and the USBFS to leave them at 2.51 or 2.50 where 2.51 was not an option.

By changing version number you are telling PC that it's using new firmware while it does not do so.
While this will work in most cases, some risk exists unless cypress keep updating their programmer code source.
Soldering 3 wires is a less risky solution, but it makes pioneer kit unusable without re-soldering resistors.

Anyway this PSoC 4 programmer code and your PSoC5 changes are progress towards AVR style cheap dev tools. And that is important to attract more devs to this great architecture.

[paulie]

what $1 chips did you get (package information) and what prototype board did you get (41xx or 42xx)?

42xx board and CY8C4245PVI-482 chips so my prototype looks exactly like yours. It seems that two of your solutions will work for super low cost programming. Reprogramming the kit board will keep cost to $4 which is well within my budget and would fit right into the purpose of this thread. Because they referred to the board as "Stamp" I initially confused it with the Parallax product so didn't appreciate this approach. Definitely worth a try.

More importantly the ability to use an STlink dongle is great news because a couple of those were purchased, one on Ebay for $5 and another on Aliexpress for $4. This last solution is by far the easiest since there are no modifications involved. Installing the Keil package on my netbook  is a bit of a pain but when the LPC and STM8 projects are done this may be next.

Thanks for taking time to explain this and hopefully more details and discussion are coming.

[paulie]

I just go to Digikey and pick what I want, and grumble if I'm "forced" to second-source stuff from Mouser.

For me it's the other way around. Mouser prices are significantly less than Digikey (although pressed they will usually match) but the clincher is inability to use Paypal at  Digikey. My account always has balance so fees are avoided and I'm very fond of the protection policy which has saved my butt more than once.

Hey, Paulie.. if you or westfw end up with an ASM file for the 812, let me know.

I have a package with linker script, source, and hex identical to the ST one on first post here but for LPC. I was planning to put up the ST8 stuff first and don't know if it will be timely or useful to you but will see if I can dig out the LPC version when i get home.

don't want to get distracted going down the rabbit hole of writing a proper test app just yet.

Too much time watching Matrix movies? Take the blue pill.

[paulie]

if $25 is acceptable for a dev board that can double as programmer for those $4 boards.

If these were to be used for serious development, even for hobby use, that would make sense. In my case 8 bit is still my main arena  and ARM is just  casual experimentation. There are so many families and variations cost would quickly add up. Already over a dozen different parts from 4 manufacturers and it ain't over yet so we would be talking big money if I adopted this mindset:

their programming dongles are $40 and a dev board is usually less than $100.

He considered that a bargain. I suppose it sounds great compared to the old days when hundreds for hardware and thousands for software were the norm but not in my case which is just for curiosity and fun.

[miguelvp]

By changing version number you are telling PC that it's using new firmware while it does not do so.
While this will work in most cases, some risk exists unless cypress keep updating their programmer code source.
Soldering 3 wires is a less risky solution, but it makes pioneer kit unusable without re-soldering resistors.

Anyway this PSoC 4 programmer code and your PSoC5 changes are progress towards AVR style cheap dev tools. And that is important to attract more devs to this great architecture.

I agree, I had it as 2.03 since most of the toolchain will just complain that is not up to date but that's fine, the only piece of software that refused to work was the bridge control panel that would not connect unless the kitprog was on the latest version. That's the reason I changed it and I can communicate with the external chip via I2C etc via the 5LP.

@paulie

I will work on a bootloadale version of the $4 programmer, and add more detail on the process.
Also I want to note that after the promotion ($1 which today is the last day) they wont go to $4-$5 a pop, regularly the one you got is $2.49 qty 1 I believe.

I'll add more details on the third party programmers but I don't have an st-link so I can't tell if that will work for sure until someone tries. You can create the full project in Creator including programming the UDBs and then export to keil and theoretically you should be able to use the supported programmers. The  application note regarding uVision only used ulink pro, but that doesn't mean the other ones won't work.

In theory you can use any micro controller to program the chip via SWD.

[miguelvp]

So I did look more into the project and it does have the bootloader already in there, weird they don't have the source for it just the hex file as a dependency.

But if I press the user button while powering it with the programmer installed it does blink and it will accept a new bootloadable program.

The first time I did try, I used the prebuild one in there and that bricked the prototype board, but I did reset it back with the pioneer.

So you have to load the 4 dollar programmer project and upgrade the components build it and then program the board.

I'll do a step by step instructions shortly.