Skip to main content
Topic: Web Platform v2? (Read 213665 times) previous topic - next topic

Web Platform v2?

Does anyone have suggestions for features or a chip for web platform v2? I was thinking about a 64pin chip that would give enough outputs to form the PCB int he shape of an arduino, so it could work with (some) shields.
Got a question? Please ask in the forum for the fastest answers.

Re: Web Platform v2?

Reply #1
More i/o pins would be nice as I ran out of them faster than I imagined I could (4 relay board, LCD).

Zero interest in any Arduino addon ability - from what I've seen most "shields" cost way more than an equivalent circuit. For example, an Arduino ethernet shield is around $US 46 (SparkFun) whereas an equivalent generic ethernet board featuring an ENC28J60 is less than $US 20 (Futurlec.com/Futurlec.com.au). This was actually the reason why I went with Microchip PICs in the first place when researching in which microcontroller infrastructure to invest (ie compiler, PCBs, addons, programmer, books, fading brain power...).

As for other features, four screw mounting holes would be nice at the cost of a bigger, and I guess slightly more expensive, PCB. On the other hand, if you're aiming at keeping the price close to the $US 40 mark including worldwide shipping, I guess I can survive with the double-sided tape approach.

Re: Web Platform v2?

Reply #2
I would go for more outputs. Also ENC28J60 chip on v1 is outdated as I know, maybe we need to change that to newer versions. I heard something about it not being produced anymore but I'm not so sure. Also I would suggest that we talk with Seeed and they find a suitable project box for it.

Re: Web Platform v2?

Reply #3
For a major speedbump consider using a ENC424J600 or ENC624J600 in combination with a processor with enough I/O to interface with the ENC424J600's parallel port instead of over SPI. I've sampled it in combination with a DSPIC33FJ128GP804 but I haven't found time to play with these, yet.

Here also no intrest in Arduino compatibility either, and another vote for mounting holes. Further: break out all free I/O.

That would be the ultimate webplatform :P


Edit: http://www.microchip.com/wwwproducts/Devices.aspx?dDocName=en542414

Re: Web Platform v2?

Reply #4
PIC24FJ256GB206 (with USB) would be interesting (plenty of RAM) so it's only 16MIPS if you plan to stay with 16-Bit.
If 32-Bit is an option I'd look into the PIC32MX with Ethernet and USB ... PIC32MX675F256H/PIC32MX675F512H/PIC32MX695F512H


Re: Web Platform v2?

Reply #6
I have a PIC32 Ethernet Starter Kit, have to write a review about it, will be finished soon. But won't suggest it for a ethernet kit for hobbyist use. I vote for a dsPIC as it is proven with the first Web platform and because it has DMA and other capabilities.

Re: Web Platform v2?

Reply #7
Um -- PIC32 does also have DMA channels.  The better ones even have up to 8 DMA channels.  I used a PIC32 on this board [1] quite a bit, and prefer it to the dsPIC.  Sure, the dsPIC has PPS, but although the PCB layout may be come out simpler and the pin layout be more flexible, I also find the additional configuration step required by the PPS mapping a source of messing up the code.

The current sourced/sinked on each pin can be as high as 24 mA, much more than the dsPIC33 can provide.  Processing speed is amazing.  Thanks to its 5-stage pipeline and prefetch cache, it can achieve up to 1.5 MIPS per MHz and can run at 80 MHz.  The better models have an Ethernet MAC already built in, so there is just an external PHY chip necessary (as done on the Microchip Ethernet Starter Kit).

[1] http://code.google.com/p/cui32/

Re: Web Platform v2?

Reply #8
Hmm, my bad, have to talk about these in my review. Thanks for pointing them out!

So only advantage of 16 bit PIC's is PPS? Wow, that's a shame! They even have Linux C32 compilers out there for it! I don't know which one is better now: to use a dsPIC or PIC32. But I would suggest putting a different MAC for each one somehow.

Re: Web Platform v2?

Reply #9
Might be nice to use a USB chip with a USB bootloader, especially once we get the open source USB stack going in full. The FTDI chip could be dropped, and maybe the price reduced even with a fancier PIC.

I would like any updated hardware to be 100% free of non open source code. It should highlight the uIP port, OS USB stack, and OS FAT SD card driver, all code distributable.

All future network dev boards should have some on-board input for demos (temperature, button). Also should have an EEPROM or MAC with unique MAC address.

A PIC32 dev board could be fun, would it make more sense to develop a cheap ARM platform instead? I have a nearly-complete design in the workshop somewhere, I'll try to remember to post it.

I worry about the availability of open source drivers for the new MAC/PHY chips. Writing drivers for new chips on unproven hardware is a lot to take on in a single project. This might be a good long-term goal.

It sounds like there could be interest in two projects -

1. A 16bit, hobbyist friendly board, an update to the web platform. Cheap as possible, provides example of minimal embedded network node with lots of resources for sloppy programmers (like me!). More IO, but still using routing and chip sizes that a hobbyist can conceive of dealing with. Maybe move to an integrated USB chip, but I'm not sure there are any ds33s with USB. This board answers the question: how do I control a LED over a network.

2. A 32bit ARM or PIC board with some serious gear. Multiple power rail requirements, external PHY, etc.

Attached a picture of a minor update of the Web Platform (no planned availability). It has 5 additional (analog) I/Os, USB/external power supply selection, and mounting holes (yeah!).

The Web Platform is kind of the ugly step child of the DP family. It's my go-to board for various live internet things (Tweet tree, graffiti wall, etc), but it doesn't have a standard, regularly updated firmware.
Got a question? Please ask in the forum for the fastest answers.

Re: Web Platform v2?

Reply #10
On board USB would be nice but dsPIC33 series doesn't have USB, for 16-bit PICs, only PIC24F series have USB.

Distributable code would be nice, otherwise people cannot share their projects freely and may stop tinkering with it (just like me!but hopefully I'll start playing with it again).

ARM is a good platform, I would support it. But are all the libraries available for it?

You are right about the new PHY chips but do we have access to the older one? If so it's not a big problem, v3 can be developed with the new PHY chip.

I think I would go for both boards for different kinds of projects. Hobbyist board for kind of a home automation system and the other one for... well, other stuff! :) By the way, embedded linux anyone?

Re: Web Platform v2?

Reply #11
Power-Over-Ethernet would be nice.
I'm not sure how it works, but there is something like 48V on one of the pairs in the rj45.
Hopefully there's not any software negotiation.
Then we could change the power regulator to accept that kind of input voltage and a diode so we dont send any current out on the network if it has another power source.

Re: Web Platform v2?

Reply #12
[quote author="honken"]Power-Over-Ethernet would be nice.
I'm not sure how it works, but there is something like 48V on one of the pairs in the rj45.
Hopefully there's not any software negotiation.
Then we could change the power regulator to accept that kind of input voltage and a diode so we dont send any current out on the network if it has another power source.[/quote]POE is very handy for remote installations that might be far from power and inconvenient for changing batteries.  If you can run ethernet, then POE gets you power, too.

However, there is not a single standard for POE.  Some flavors of ethernet use all pins, and thus are incompatible with POE.  Where POE is possible, there is more than one way to put the power on the spare wires, and thus you run the risk of burning things up if you plug in POE without first confirming compatibility.

The reason I mention this is that POE could be a liability in a DP product aimed at novice electronics hackers.  Maybe the best solution is to etch support in the PCB for POE, but not populate the parts.  It might even be possible to support multiple POE wiring conventions this way.

P.S.  If there were software negotiation, then it might solve the problem of multiple, incompatible conventions.  But the only way for software negotiation to work is if the remote board has rechargeable battery backup - otherwise it would not have the power to negotiate in the first place.

Re: Web Platform v2?

Reply #13
[quote author="ian"]
It sounds like there could be interest in two projects -
[/quote]

Do a quick price check of how much more expensive a pic32mx platform would be compared to dsPIC platform. If I'm not mistaken, difference is minimal ... for e.g.

PIC32MX695F512H (64pins,     512K Flash (plus 12K boot Flash), 128K RAM (can execute from RAM), 8 Channel General Hardware DMA Controller, Flash prefetch module with 256 Byte cache... 4 Dedicated DMA Channels for USB OTG & Ethernet ) is 7GBP @farnell (4.5 if you get more) ..

dsPIC33FJ128GP204 that web platform currently uses is 4GBP (or 3 if you get a lot of them) and it is 16bit 44pin mcu without even usb not to mention ethernet ..

If you just add a price of FTDI on the dsPIC the pic32mx is cheaper solution ... ok you might go with some dsPIC24 like for e.g. PIC24FJ256DA206 but then you are again on the no ethernet, 16bit, quarter the ram/flash solution that cost 0.1-0.2GBP less then pic32MX solution..

I just don't see a reason to go 16bit, 32bit is same price and much more power + additional hardware. The only benefit of the dsPIC is the "move the pins around", it is cool feature but I prefer power :D

Re: Web Platform v2?

Reply #14
Quote
the only way for software negotiation to work is if the remote board has rechargeable battery backup - otherwise it would not have the power to negotiate in the first place

Your right of course, wasn't following through on my own train of thought.

Even if there's several ad-hoc solutions to PoE, why wouldn't we adhere to a standard, ieee 802.3af?
The problem with which pair of pairs is carrying the supply could be solved with two graetz bridges (or eight diodes, or 16 to be absolutely sure even if someone is using a single pair to carry the power, we're not aiming for Gb ethernet that uses all four pairs).
The real problem I see is that we could not use the connector with internal magnetics as in v1 and the wide input range power regulator (48V wall warts are hard to come by so input should be something like 9-60V tolerant).
For PoE to work we need access to the centre tap of the primary side of the ethernet transformers. That we don't have with previous connector.
But if that connector isn't available for v2 (it's gone from seeds depot) I would find it a nice, not need, feature. Imagine only the one cord going into your black box of TCP/IP awesomeness.