Taking Gen 8 wishlist

Hi everyone.
Some background. I am an EE with about 15 years professional experience.
I am busy putting together an ARM all DIP system. If the community wants I will release it as Gen8.

Some features so far.
1x Low cost DIP LPC1114FN28/102 AMR processor.
2x low cost high speed high current PWM outputs.
6x TB6560AHQ 3A stepper drivers. available on ebay for US3 each.
USB via 8 pin DIP AVR.
All (so far) on a 10cm x 10cm double sided PCB.
3 limit switch inputs.

Questions
Thermocouples or thermistors?
What else would be good to have on board? i/o? LCD, buttons etc.
A piezo would be nice, easy electrically but would push us out of the 10cm x 10cm boundaries.

Murray.

The PCB so far.
Just to give an idea of the form factor.

Cool idea. Deffinitely release it. I wouldn't call it Gen8 if I were you. We're trying to get away from the GenX naming system since there really aren't any official electronics anymore. I would love to hear what results you get with the TB6560's I've done some work on Gen7T and am running the contest [reprap.org]

For me I would say keep the 10x10 form factor. That way it's easy for a few friends to have some made at iteadstudio, etc. for cheap. Of course other benefits may override that prefference.

I think LCD is an add on for most people right now though that may be changing, so I would think that compatibility with one of the add on boards around here would be a plus but not necessary to have it built in.

With only 28 pins, are you running out of IO?

Also I'd love to get your oppinion on my similar project when I get it routed.

re thermocouple vs thermistor. I think thermistor is currently the default for most electronics and firmware. It might be nice to allow for both but I would probably go with simple and cheap. That's just me.

I would love to use the DIP LPC1114FN28/102 for a Rostock-style printer, as it need more power to process the delta coordinate system, and the design can handle higher speeds due to the light-weight printhead. Four stepper driver would be needed with this design. I would also like to see a design with just SD card printing, so we could do without a computer interface, once we had a programmed chip.

Most ARM32 chip need a board for surface mount soldering, but with a DIP chip, I would rather use perf board, to keep it simple and cheap to build at home. So I will be following your design with most interest!

OK, you convinced me. Gen7 will get an ARM, too. On a single sided board, of course.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

A schematic I've got an urge to pull out a breadboard

Schematic is in the making.
There is a lot of work still to be done on the ARM port. For anyone just wanting to build something and not having to mod and hack, I'd recommend building a Gen 7.

Traumflug, hephaestus perhaps we should work together.

My aim is to build a builder friendly (ie DIP) ARM system that scales up.
I am happy to use double sided boards, because I don't want to make my own and seeedstudio does it pretty cheaply.
Traumflug wants to make a system that can be built on a single sided board.
hephaestus wants a system that can be built on on strip-board.

So the common parts are the DIP parts and the ARM processor as there is only one DIP ARM processor out there at the moment.

I'd like to work with the TB6560 because it is DIP, high current, cheap (US3) , can be driven with upto 36V and 3A. Although the CPU side should be able interface with most clock and direction driver chips.
Here is a copy of the preliminary schematic I was working on, although some things have changed since then.
I am using the 74hc595 shift registers to drive the CLK and DIR pins on the drivers. The processor will output upto a 16 bit word at a time so 2 can be daisy chained with automatic loading after the 2 bytes has been clocked out. The drawback with this approach is that you reduce the maximum speed you can change the pins to the cycle time of the SPI port. A quick glance suggests a max update speed of about 25MHz /18 = 1,4MHz. more than enough. And enough pins to drive 8 stepper drives.

The processor runs on 3,3V. This means driving the MOSFETS (and other i/o like LCDs) directly becomes a problem. Ive connected a 556 timer as a mosfet driver. The device acts as a voltage translator driving 12V to the output and with upto 200mA high and low, making high speed switching possible. The 556 is also very cheap and available.

Not so much I want a breadboard, but it makes a nice circuit testing environment prototyping on the breadboard makes it quick to test and get onto the software side...

I've said it before, I want a cortex AIO, with user replaceable drivers... so I don't brick a full board with a driver issue... I was starting to play with a drawing to make a cortex a drop in replacement for an atmega644/1284

Probably other ARM electronics have a solution for the MOSFETs already. I think there are types which work with 3.3V on the gate.

Using the RS-232 protocol between the ATtiny and the ARM won't work, as the ATtiny45/85 doesn't have a hardware RS-232 register. Software RS-232 is doable, but limited to 4800 baud. I'm currently trying to figure out how the ATtiny can be programmed its self as well as serving as a programmer for the ATmega (ARM here) and a SPI-serial device for regular communication. Might be doable with a jumper on the reset pin.

Regarding collaboration: well, obviously we all need a firmware. Other ARM electronics have all their own firmware so far, doubling a lot of work. This isn't neccessary.

How are these ARMs programmed? Hopefully not on an external programmer

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Traumflug Wrote:

> How are these ARMs programmed? Hopefully not on an
> external programmer

Its pretty well explained at [hackaday.com]

Edited 1 time(s). Last edit at 10/19/2012 12:00AM by hephaestus.

As you may know, I am a fan of the ARM chips, and I think its great there is now an ARM in a DIP package, the NXP chips are also easy to use and well supported.

My main worry here is that the LPC1114 is only 32KB, I think it is going to be a *very* tight squeeze to a get a decent firmware in there. The ARM firmware I am porting for sango-bc comes in at least 60KB, I will be going to a 128KB chip for the next version. Other ARM platforms are using 256KB or 512KB.

Its going to require some lean code, there's no doubt, but I've got a bright shiny nickle that says its doable...

Nuvoton has a couple LQFP package units, that are interesting, given they appear to be 5v tolerant, at least by the data sheet / vdd...

Dip samples are here. Not a skinny dip like the ATmega8 rather a wide DIP like the old EPROMS.
Now to set up the toolchain. I will put together a tutorial as I go.

The big question is can we fit the firmware for deltabots onto this chip, considering the extra ram requirements for the delta calculations. The modified Marlin and Repetier firmware for the Rostock don't seem to function on Sanguinololo (4K ram).

4K RAM!
Ive not looked at the code but that sounds like a lot more than needed. With a bit of effort I'm sure we can fit it in.

I am wondering how this is progressing? I am interested in getting a Teacup port to LPC1114, I built Teacup for AVR and it came to around 17KB which is very encouraging.

I now have a couple of LPC1114 dev boards, I have run the obligatory blink project If you have any code I could look at, or want me to help with any problems, let me know.

bobc, do these dev boards come with a schematics? It shouldn't be too difficult to slap such an LPC into a Gen7. Everything else there is working already, the USB adapter starts to be useful.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

I think it could be even better making a small pcb with a lpc1343 just to plug on the atmega socket of the current Gen7 board. This ARM is almost 99% register compatible with the lpc1114 and it have an onboard USB controller with CDC, HID and mass storage (no software stack needed).
It's only smd but it could be an easy way to start developing with ARM without remaking the whole Gen7.

Sure, the two boards I have are these : a DIY Open Hardware board designed by Opendous, [github.com] Despite the name it also takes LPC111x chips (the USB is not used ).

A cheap one I found on ebay: [www.micro4you.com]
I think Olimex sell a cheaper module still.

I can't seem to find the DIP versions of LPC1114, but I believe it is near identical to the LQFP chips I have. I like the idea of making a module in a 2.0"x0.6" package to fit the AVR socket, I think that a similar schematic to the one Adrian has for the MBed could be used.

I will make a start on porting some Teacup code soon.

Remorilla Wrote:
-------------------------------------------------------
> I think it could be even better making a small pcb
> with a lpc1343 just to plug on the atmega socket
> of the current Gen7 board. This ARM is almost 99%
> register compatible with the lpc1114 and it have
> an onboard USB controller with CDC, HID and mass
> storage (no software stack needed).
> It's only smd but it could be an easy way to start
> developing with ARM without remaking the whole
> Gen7.

Something like this [squonk42.github.com] or this [micropendousx.googlecode.com] perhaps?

Squishing those boards down should be possible, might take some nifty layout. However, if there is clearance on the Gen7 board, then the module could extend beyond the DIP footprint, which might help. The LPC1114 and LPC1343 are pin compatible apart from the USB pins, which is handy.

There are open source libraries for both lpc1114 and lpc1343, but with BSD license.

Hi all, I'm starting with ARM chips, and I'm planning to start a Rostock 3d printer. I have a STM32F4DISCOVERY board. It's a powerful stm32 cortexM4 chip inside a very cheap development board (15EUR!!!) with a USB programmer. Maybe could be a good idea to develop a shield for this board (RAMPS like). This could be the cheapest and powerful reprap electronics board!! ><

@Chele69: I've been down that road. The conclusion I've come to is this: The STM32F4DISCOVERY is not suitable for mainstream reprap. It's absolutely a great board, it can do the job, no problem, but it's not going to fit the requirements of the mainstream reprap electronics.

As I understand it, the requirements are these:
1) No proprietary (non-opensource) boards
2) Hobbyist friendly (no SMD components, single-sided)

Gen7 still violates 2), in that it has the pololu stepper drivers, but from what I understand, there is active development in the works to remove those. Gen7T is an example of this.

RAMPS also violates 2), but it does so even more. It uses the pololu drivers like Gen7, but also an Arduino Mega. RAMPS also uses a shield which is effectively just a header board (it does have a few components, but not many) which drives up the cost. Gen7 contains everything you need except the stepper drivers, which brings the cost down. Gen7T brings it down further by including the stepper drivers too.


The problem with the STM32F4DISCOVERY is that it violates both 1) and 2). It's not that it isn't a great board, it's that it's not quite right for mainstream reprap adoption. It also has a bizarre license (though I suspect that first-sale doctrine makes that irrelevant, at least in the U.S.).

That being said, go see if you can make an awesome reprap controller. If it works out, I suggest you design a replacement for the STM32F4DISCOVERY, do as much cost containment as you can on it, and open-source it.

I would also suggest you take a look at R2C2, which might be similar to what you're designing.

Quote

2) Hobbyist friendly (no SMD components, single-sided)

This is a controller people on IRC are pretty excited about:

[dustsreprap.blogspot.de]

All SMD, super fine pin spacing, wasteful feature count, not solderable by mere humans. You see, people's opinions about what a good controller is, varies widely.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Traumflug Wrote:
-------------------------------------------------------
> All SMD, super fine pin spacing, wasteful feature
> count, not solderable by mere humans. You see,
> people's opinions about what a good controller is,
> varies widely.

Oh, naturally, but on this forum, it has to be DIP. For example, when I proposed a new generation of electronics based on the LPC176x series devices, a significant majority of respondents was not interested in a new generation of reprap electronics if they weren't 100% DIP. I can find you the thread and the quotes if you want, but I don't think that's necessary, since you were active in that discussion.