New AVR board feasibility study

Go To Last Post
33 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Halo guys,

 

I apologize for off-topic, but I have question here and would appreciate one minute of your time here :) 

 

we all are using Arduino based devices for our project. But we want to get serious about deployed hardware and design specific 
CPU board for our Control/Monitoring/PLC projects. Goal is to develop reliable, industrial grade board with possibility of programming 
in Arduino IDE as well as in Atmel IDE. So there will be bootloader and JTAG interface for sure. 

As we are planning to do it Open source later on the road, I would like to know some insights/wishes from you regarding hardware and
'on board' capabilities in terms of Processor (ATmega2560, SAMx), Communication interfaces (Wifi, BT, Ethernet, RS232/485, CANbus, 1-Wire, 
USB, GSM, etc.), Digital and Analog I/O interfaces, Oled display, Buttons, LED indicators, Real time (RTC), SDcard, EEPROM, FLASH, Environmental
sensor, and so on...)

 

One question is, if there is preference to develop 'lean' module with CPU and supported circuitry  only and communication and I/O connectors 
on separate base board where CPU modlule will be 'clicked' or to design few boards,  with specific all-in-one configuration on board. 

 

Second question is, if design of the board should match any standard DIN rail enclosure.  

 

We are still not decided which way to go, therefore some advice from you guys is needed as you could be potential users of the HW too :)

 

So please give it a though 

 

Best regards

Tomas

Last Edited: Sun. May 12, 2019 - 05:00 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Not entirely seeing how this will be "better" than Arduino already is?

 

What is your USP?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well, let's see... It should have HDMI output, a couple meg of program memory, a meg or two of ram, wifi, usb, a gigahertz or two, bluetooth, an FPGA and cost about $2 each from Amazon.

The largest known prime number: 282589933-1

It's easy to stop breaking the 10th commandment! Break the 8th instead. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

pecook wrote:

...with specific all-in-one configuration on board. 

 

No-one will ever agree on what that configuration needs to be. Even a simple question like 'how many serial ports' will get the answer 'it depends'.

#1 This forum helps those that help themselves

#2 All grounds are not created equal

#3 How have you proved that your chip is running at xxMHz?

#4 "If you think you need floating point to solve the problem then you don't understand the problem. If you really do need floating point then you have a problem you do not understand." - Heater's ex-boss

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"reliable, industrial grade board".

Well, that's hard...

In the industrial "field", there are EMI/EFT/ESD monsters. Your board/MCU can get upset very easily.

There exists very simple things, namely 24V relays (some big ones, 10A minimum). You just have to control such a relay. Simple. Or not?

You can use standard ULN2003/2803 drivers with no problem. They even have protection diodes internally. But I tell you: they are useless.

Unless you really put protection diodes JUST BY the relay coil, as close as possible! And the protection diodes are - guess what - not the 1N4000 type (0.7V) but Zener diodes, usually 33V-36V. There is an entire theory of protecting the relay coil... Well, that's easy to say, sometimes very hard to do. You maybe never get access to that cabinet, and there may be no one good enough to just put the diodes.

And yes, the relay coil gets you in trouble, not the contacts! 

And you just need one such relay, well, a "bad" one (not all relays are created equal, but usually chinese ones - veeeery cheap - are the worst).

I used to have one board with some relay sockets, and one or two "bad" relays. I used it extensively to test a lot of MCU's about their immunity to noise. But now I have two boards, the new one is worse smiley

 

With the old one, almost every microcontroller functioning on the INTERNAL OSCILLATORS is immune. Many of them with external crystal AREN'T.

 

With the new one, almost NONE are immune, even on their internal oscillators. The only one family by now which is immune is the Nuvoton (8051 and Cortex) which by the way is kind of a "Wow", because they are immune even with external crystals. They are specified with 8kV ESD/ 4kV EFT which seems just real, not marketing. I'm waiting for my Kinetis E-series prototype for this board, very curious about it. AVR, STM8, STM32 are just useless with this board. 

 

So, if I would recommend something for your "reliable, industrial grade board" I would recommend:

- Nuvoton (almost any microcontroller, if they say 8kV/4kV ESD/EFT - they are good)

- Kinetis E-family.

Sorry, probably no Arduino, no Atmel-Studio...

 

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Who is going to use this "generic"  processor board...why would they want yours, instead of a known PLC?   

 

Perhaps if they are designing custom equipment (say, a new industrial lathe)...if someone was creating a new line of arc welders, they certainly be designing some circuit boards the go into it  (which could include the micro, control panel, switches, displays, etc).

It seems for higher-end equipment, a custom board would probably be created, tuned to the exact needs.

 

I f you create any generic board, be prepared for china to duplicate it , with free shipping, for $5

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Halo guys,

thank you for answers and insights. I will try to summarize reply in one message and clarify purpose of the hardware bit more.

Yes, market is certainly saturated with zillions of Arduino shields, boards and everybody needs different boards for different purposes, right. If you will need some functionality, you need shield for that, then maybe another one if other functionality is needed, and so on. In the end you will end up with bunch of ugly boards stacked on top of each other to fulfill functionality. This is case of average user/hobbyist using Arduino platform for own purposes. And do projects using multiple shields is frankly a bit of pain.

So idea here was to combine multiple 'shields' into one board or 'shield'. Say Ethernet/RS232/RS485/Canbus/bunch of protected, isolated I/O to address needs of folks trying to do various home/control projects, for example. And this board/shield would contain components related only to 'control' with Oled display too for example. But we are still talking about average "buy this board in eshop' customer. Is for peoople who insist to use Arduino and have no idea about IEC61131 and related PLC hardware.

We will not be targeting industrial customers or projects. Arduino platform (Controllino and other home brews) are not suited for that job at all, as there is limited complexity and other major issues. For those projects we use ABB or Siemens equipment. We just want to produce combined industrial grade board/shield to make life easier and cheaper for average user/hobbyist. Meaning combined functionality, ESD protected, EMI/EMC friendly, etc. Just 'Get & Deploy'. Hardware design is not problem, We are doing this every day :)

We got questions from other forum so there is answer to them:

what return will the business get from all of the development time and cost
Money :) And we are exploring here if the idea is relevant and worth of the effort or not

What customers have you in mind.
Average Arduino users/Hobbyist/Students/Geeks

What market research have you undertaken to see what products those customers might want?
So far this is first post discussing whole idea and its validity

How many potential customers do you anticipate for the board you are thinking of?
Say 100/200 peoples would be fine to pay for development cost by buying final product

What percentage of the potential customers might buy your board?
Well, this is Million dollar question. But I believe this should be answered by Arduino equipment e-shop seller. Because thew will know what kind of boards/shields are sold most. Therefore creating 'one-in-all' board/shield combining all most needed features would be desirable and would address needs of major group of the potential users.

Why would they buy your board?uy 
Because they are consumers, usually doing stuff for own use. Therefore cheapest, multiple functionality & $$$ friendly board/shield will be 'right on the spot'. Multiple, combined shields are expensive in the end.  Lets say that you will need Ethernet, RS485, LTE/GSM,  8x8 I/O. You will easily end up 120USD for those 4 shields, to create 'Eiffel tower' of functionality.  Or you can consider 100USD board/shield covering all of this plus some extras like RTC clock, SD card. 

So idea is to get rid of multiple shield approach and create one comprehensive piece of hardware instead. Which would be neat and cheaper. And question is 'what to combine ..." and if is worth. China guy can copy this design as it will be open source later on. I don't care about that.  

Best regards
Tomas   
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

 

You might be better off making 'hardened' add-on IO modules in an already existing format eg Mikroe.

 

...along with a more universal CPU board.

 

#1 This forum helps those that help themselves

#2 All grounds are not created equal

#3 How have you proved that your chip is running at xxMHz?

#4 "If you think you need floating point to solve the problem then you don't understand the problem. If you really do need floating point then you have a problem you do not understand." - Heater's ex-boss

Last Edited: Sun. May 12, 2019 - 07:13 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

pecook wrote:
Goal is to develop reliable, industrial grade board with possibility of programming 
in Arduino IDE as well as in Atmel IDE. So there will be bootloader and JTAG interface for sure. 

 

I use PLC's frequently and I can tell you right now I am not going to switch from known platforms that have software suites that are graphically based to a board that I now have to write everything at the root level, and then debug hundreds of lines of code should there be an issue.

 

pecook wrote:
As we are planning to do it Open source later on the road,

The industrial community is not so keen on Open Source.  Just keep this in mind for your target audience.

 

pecook wrote:
We just want to produce combined industrial grade board/shield to make life easier and cheaper for average user/hobbyist. Meaning combined functionality, ESD protected, EMI/EMC friendly, etc. Just 'Get & Deploy'. Hardware design is not problem, We are doing this every day :)

 

Take a look at this:

https://www.automationdirect.com...(stackable_micro_brick)/plc_units/c0-00dd1-d

 

$69.00 for an expandable PLC that has I/O on board.  Software is free, the programming cable is $18.00.  I use these a lot in Security applications.  Elsewhere I use Allen Bradley/Modicon/Omron/Siemens.  Point is, that if a hobbyist really needs a PLC, theres already a "been there, done that".  How does your board only, root level programming design become more cost effective?

 

AM I telling you not to pursue your dream?  Absolutely not.  I am only showing you, that based on your two posts that you have a lot to compete against.

 

JIm

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

One of the "clever" things about Arduino and shields is that you can put together whatever combination you need without paying for bits that you don't. Unless you can make your board down in the cents for each function provided then people are going to perceive it as paying dollars for bits they don't need.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you want to make money---develop a great board for a niche app that is not well served....Here is our universal fingerprint analyzer board that can be configured  like this , like that ,has these configurable options built in, etc/...maybe someone will buy a bunch to make/sell some detective equipment.

Otherwise, it's tough since you are trying to design a board as a component for a completely unknown purpose ("industrial"--is a market, not a purpose).

I think some freak was making motorcycle engine monitor/control boards....something like that generates interest from a sizable garage audience, who are quick to buy & peel out. 

 

Remember Big Billy Bass Mouth?

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

As explained, targeted audience are students/hobbyist/geeks trying to use Arduino platform for control/monitoring purposes due to programming convenience. 

 

Nobody from 'real automation world' will need or use this hardware, for sure :)  Idea was to utilize already available CPU boards (Arduino, STM32, ESP32, etc.)

and just do 'shield' or extra board with some neat, protected 16DI/16DO/8AI/4AO/Comunications in one 'package' to experiment with.

 

Yes, todays PLC's are dirt cheap :) I do all my projects in CODESYS based PLC's like ABB or Beckhoff. But for average user/hobbyist is commercial PLC unknown world.

 

T.   

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes, its about to find community who needs some specific HW and design it for them....  What is ' Big Billy Bass Mouth' ?

 

T.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

rammon wrote:
- Nuvoton (almost any microcontroller, if they say 8kV/4kV ESD/EFT - they are good)
Another is Silicon Labs EFM8.

AVR don't have an ESD spec; likely none for PIC.

AVR can survive and be operational though do require external circuits, limiting of ESD/EFT/lightning current loop's area and location, and a correct layout.

 

https://www.avrfreaks.net/forum/i-posted-new-version-my-crc-tool-crcsntool-projects-area#comment-2666626

...

 

There are currently about 28,000 lines of code in the controller (ATMega1284) but one of the biggest challenges is lightning:

[URL]

...

AVR040: EMC Design Considerations

4. Design Rules

(page 11)

4.5. ESD and Transients

EFM8UB3 Data Sheet | Silicon Labs

(page 1)

...  8 kV ESD protection, ...

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

pecook wrote:
Money :) 
https://www.azquotes.com/quote/563356

Instead, would you consider revenue?

as that's what an accountant will measure via the ledger.

pecook wrote:
And we are exploring here if the idea is relevant and worth of the effort or not
Recommend complete, precise, and correct risk analysis (risk = probability of failure * consequence of failure, then, evaluation of risk); the flip side of risk is benefit.

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Just Go right ahead, don't let any negativity stop you.
I would give it a try sure others would too.

Who is "We"?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

kerlin wrote:
I would give it a try sure others would too.
That's a bit like betting on horses or shares. Only ever invest money you can happily afford to lose.

 

Sorry if this is a rather pragmatic approach to life but it helps to keep bread on the table :-)

 

PS of course the other side of the shares/horses/lottery thing is that it *may* pay out big time if you strike lucky

Last Edited: Mon. May 13, 2019 - 03:38 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes, lighting is b**ch :) But proper ESD measures can eliminate that. Thank you for links, will take time to study....

 

T.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

kerlin wrote:
Just Go right ahead, don't let any negativity stop you. I would give it a try sure others would too. Who is "We"?

 

Hehe, thank you for heads up. Project will definitely happen. Question is which way we will go. What king of device will be designed.

We i Me with couple of geeky electronics friends..

 

T.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

clawson wrote:

kerlin wrote:
I would give it a try sure others would too.
That's a bit like betting on horses or shares. Only ever invest money you can happily afford to lose.

 

Sorry if this is a rather pragmatic approach to life but it helps to keep bread on the table :-)

 

PS of course the other side of the shares/horses/lottery thing is that it *may* pay out big time if you strike lucky

 

Yes, fully agree. I am pragmatic person too and time is precious. And to eliminate useless horse (to use same analogy) we are searching barns for good one :) 

Let's see ....

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

pecook wrote:
And to eliminate useless horse ...
no such

Grandson of Secretariat, JD blows out his birthday candles! - YouTube (26s)

Buck (2011) - IMDb

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I would consider moving to this "Blue Pill" board:  https://www.ebay.com/itm/STM32F1...

 

It is somewhat cheaper than the high-end AVR ICs with the basic same peripherals and 72MHz system clock, 128K Flash, 20K SRAM.  It is Arduino compatable, but doesn't come Arduino-configured from the eBay distributor. It has real-time clock and USB built-in.

 

Instead of designing a CPU board, it would be easier to design a series of daugher-boards that plug into one or both rows of the 0.1" standard header plated-thru holes on the Blue Pill board.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

pecook wrote:

As explained, targeted audience are students/hobbyist/geeks trying to use Arduino platform for control/monitoring purposes due to programming convenience. 

 

Nobody from 'real automation world' will need or use this hardware, for sure :)  Idea was to utilize already available CPU boards (Arduino, STM32, ESP32, etc.)

and just do 'shield' or extra board with some neat, protected 16DI/16DO/8AI/4AO/Comunications in one 'package' to experiment with.

 

Yes, todays PLC's are dirt cheap :) I do all my projects in CODESYS based PLC's like ABB or Beckhoff. But for average user/hobbyist is commercial PLC unknown world.

 

It makes sense to follow pin-out standards, Arduino is one, and Raspberry Pi 40 pin header is another. It should be possible to support both ?

You need to first define, exactly what you (your customers) need, on that IO board.

 

Nuvoton (already mentioned) have a number of standard modules that have Ardunio pinouts, 

https://direct.nuvoton.com/en/tools-for-microcontroller/

Their 5V ARMs would be a logic choice as that can match any Ardunio, and give you some horsepower extras, especially in the 5V M4 cores.

 

Raspberry Pi has some LCD Display suppliers claiming 128MHz SPI links, via a CPLD, so you could look into a CPLD based IO expander, that can use all that available speed.

A CPLD would allow you to include some quadrature counters, which are likely to be popular.

This could be a good starting point :

http://www.latticesemi.com/Products/DevelopmentBoardsAndKits/iCE40UltraPlusBreakoutBoard

 

If you need DACs,  choose a small MCU with DACs.

Microchip have attiny's with up to 3 x 8b DACs, or  parts like EFM8BB3 have 2 or 4 12b DACs, and the EFM8LB1 has 14b ADC and 12b DACs, allowing you to offer quite high performance Analog expansion.

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

One question is, if there is preference to develop 'lean' module with CPU and supported circuitry  only and communication and I/O connectors 
on separate base board where CPU modlule will be 'clicked' or to design few boards

Connectors are a weak point, especially if you are aiming for "hardened boards"...putting the processor on a separate connector-needy board from the relays, DAC, motor drive, etc saves neither space, nor cost.  Put it all on one board, or just sell this as an add-on to an rpi

, nano, other other cheapo board...but then that wouldn't be considered much of a "hardened" system---those boards are pretty weak. 

 

If you want to make hardened systems then make hardened systems--accept no weak links.

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

avrcandies wrote:

Connectors are a weak point, especially if you are aiming for "hardened boards"...putting the processor on a separate connector-needy board from the relays, DAC, motor drive, etc saves neither space, nor cost.  Put it all on one board, or just sell this as an add-on to an rpi

, nano, other other cheapo board...but then that wouldn't be considered much of a "hardened" system---those boards are pretty weak. 

 

If you want to make hardened systems then make hardened systems--accept no weak links.

 

Fair point, but the OP has stated this :

As explained, targeted audience are students/hobbyist/geeks trying to use Arduino platform for control/monitoring purposes due to programming convenience. 

 

ie students/hobbyist/geeks are not really "hardened" system customers.

 

An alternate approach to 'hardened systems' is to physically separate the main CPU from the IO cables, and use opto isolated & SSR  etc IO modules.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

 

Fair point, but the OP has stated this :

As explained, targeted audience are students/hobbyist/geeks trying to use Arduino platform for control/monitoring purposes due to programming convenience.

ie students/hobbyist/geeks are not really "hardened" system customers.  

 

But we want to get serious about deployed hardware and design specific CPU board for our Control/Monitoring/PLC projects. Goal is to develop reliable, industrial grade board 

We just want to produce combined industrial grade board/shield to make life easier and cheaper for average user/hobbyist.

Good point; to avoid confusion they should simply advertise they are making hobbyist shields that are protected against ESD, reverse polarity, short circuits, overtemperature and other gory goofs.

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Who-me wrote:
A CPLD would allow you to include some quadrature counters, which are likely to be popular.
Likewise by FPGA.

Quadrature - Xcelerator Block | Alorium Technology

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi, yes, STM32 is another chip of our interest. We have been exploring idea to do something around this area too :)

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

avrcandies wrote:

 

Connectors are a weak point, especially if you are aiming for "hardened boards"...putting the processor on a separate connector-needy board from the relays, DAC, motor drive, etc saves neither space, nor cost.  Put it all on one board, or just sell this as an add-on to an rpi

, nano, other other cheapo board...but then that wouldn't be considered much of a "hardened" system---those boards are pretty weak.  If you want to make hardened systems then make hardened systems--accept no weak links.

 

Yeah, whole system has to be robust, if one part will be weak then whole system will be weak.... And not cheap as ordinary boards/hardware is. 

And as pointed out, there will be probably no need for targeted audience to have such hardware. Thinking it around, I have one device in mind, with super tiny footprint yet pin compatible with Arduino board

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Something like these guys:
https://www.rugged-circuits.com/products

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Who-me wrote:

 

It makes sense to follow pin-out standards, Arduino is one, and Raspberry Pi 40 pin header is another. It should be possible to support both ?

You need to first define, exactly what you (your customers) need, on that IO board.

 

Nuvoton (already mentioned) have a number of standard modules that have Ardunio pinouts, 

https://direct.nuvoton.com/en/tools-for-microcontroller/

Their 5V ARMs would be a logic choice as that can match any Ardunio, and give you some horsepower extras, especially in the 5V M4 cores.

 

Raspberry Pi has some LCD Display suppliers claiming 128MHz SPI links, via a CPLD, so you could look into a CPLD based IO expander, that can use all that available speed.

A CPLD would allow you to include some quadrature counters, which are likely to be popular.

This could be a good starting point :

http://www.latticesemi.com/Products/DevelopmentBoardsAndKits/iCE40UltraPlusBreakoutBoard

 

If you need DACs,  choose a small MCU with DACs.

Microchip have attiny's with up to 3 x 8b DACs, or  parts like EFM8BB3 have 2 or 4 12b DACs, and the EFM8LB1 has 14b ADC and 12b DACs, allowing you to offer quite high performance Analog expansion.

 

 

 

Thank you for links and hardware tips. For commercial success one has to follow pin-out if possible and it would be definitely done that way.

 

Regarding CPLD based boards. They could be very powerful, but I think not quite useful for broad spectrum of the users. QEncoder could be 

connected by different way to RPI, BBB, etc.  But its interesting idea. ADC/DAC could be done by cheap SPI/I2C chips and yes, Microchip is my favorite part supplier :) 

 

I do believe that I/O (analog and/or digital) expansion will be useful for the users to play with. It could be snap-on board or detached design connected with base board

via ribbon cable for example. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman wrote:
Something like these guys: https://www.rugged-circuits.com/...

 

Already got this web recently :)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:

pecook wrote:
And to eliminate useless horse ...
no such

Grandson of Secretariat, JD blows out his birthday candles! - YouTube (26s)

Buck (2011) - IMDb

 

 

let's rephrase it, to eliminate obvious useless horse.. :)