Zikzak - crummiest 8bit console/computer ever .. but I'm making it!

[Binky]

A bit like a square signal is more squarish than square when you look at it with an oscilloscope :huh:

Not if your kit can operate at infinitely high frequencies

 

[skeezix]

skeezix, on 21 May 2013 - 05:07 AM, said: So, yeah.. do it, jump in, its awesome; its _rediculously_ fun; some money down the drain, but not too much

I just talked about that to my girlfriend and she answered.

My Girlfriend said: We don't have enough space for an electronic shed here.

(she might be afraid to have my old and large analog oscilloscope back from her parent's attic to our tiny living room )

On a more serious note, I've already got a bunch of micro-controllers (PIC and MSP430), resistors, leds, led drivers, breadboards, etc.


I'm not really afraid of burning chips (I've got most of them using the "sample button" on chips makers websites).

Spoiler

I designed a really simple circuit board that allows to use arcade controls on a SNES without butchering an original joypad (a guy told me that it works fine with snes but not with super famicom, they probably swapped some lines). To be honest, my work was veeery small, the schematics were found on the web and I found the component used in NES and SNES joypads (they used 8bit parallel to serial converters, simple and effective design imho). So I only had to choose the size of the board, place the components and route the lines…


My plan was to get it to talk to a micro-controller. At first, using the micro-controller to power leds corresponding to the button pressed. Then start making simple little games using leds, eventually little lcd screens. Slowly adding parts to get a nice little gaming platform…


But life got in the way

Spoiler

I think that my main problem is that I can deal with the logic using "ideal logic gates", but I'm a bit lost with the reality of electronic. A bit like a square signal is more squarish than square when you look at it with an oscilloscope :huh:


EDIT : I'm dusting my "electronic parts" box, looks like I forgot to bring my capacitors with me when I moved… (there were only a bunch of tantalum ones) But my MSP430 launchpad is still working perfectly.

I have a big house, but its all visible and open; my basement is fully finished.. so no storage, no workbench; I'm the sort of person who really really needs a workbench around, for all the junk I may have to build somethign into the furnace-room, that tiny little area beside the furnace

Anyway, yeah, so I'm using just my computer desk; throw the anti-static mat on top of it (the mat is too big, but I let the far-side near the monitor curl up, it covers some of the cable mess ); the anti-stat mat is like 2 or 2.5 foot square, and thats my work area; I picked up a Vellemen lamp .. somethign like this:

http://www.ebay.com/itm/White-Velleman-12W-Desk-Lamp-with-Magnifying-Glass-3-12-Dioptre-/271184807060?pt=Lamps_US&hash=item3f23dfc494

Solder station goes beside it, beside my laptop.

The rest is a _pile_ of parts; I've got a 'briefcase' style toolchest, and its full up with bits.. needlenose, flux, solder, arcade buttons, clamps, etc etc; I used to carry it aroudn to do arcade machine repairs, but now I've fille dit up with a pile of other stuff.. 'helping hands' 'third hand' style little guy, a couple magnifiers to see better, spools of wire, etc. And craft boxes (you know those boxes with little 1-2" slots in it, for storing individual bits of things? like a fish tackle box? So I've got a stack of those .. sort of worried about static killing ICs, so trying to keep ICs wrapped in the baggies they come in and stuff them in there..

.. but it doesn't take up _that_ much space; its not attractive, to be sure, but its all on top of and under my desk

I never got around to doing the snes-controller-hack thing, but often wanted to Theres a lot of crazy little hacks around the house one can do..

Theres a lot of steep curves - using a CAD?! basic 'cheats' you alwaysdo, rather than know how it works, etc. .. but thery're not _tall_ hills; just a half dozen steep thigns to bang your head off, and then it starts clicking; I'm no good at it, but I got over those first humps. (And into the second, third, fourth, 100th humps now )

jeff

 

[skeezix]

A bit like a square signal is more squarish than square when you look at it with an oscilloscope :huh:

Not if your kit can operate at infinitely high frequencies

oh and.. yeah, the electronics is real time; only the _Sensing_ part, the consumer, has to worry about the malformed shapes, and most of that is taken care of with timing.. either magicly, or in your code. ie: its true, you can get really tight square wave, or really 'rounded' ones; they're all rounded when you zoom in more, but its 'how rounded' Add in temperature variation, capacitors, breadboard, bad solder joints made in a hurry etc, and it gets fuzzy; or load form variosu components doing their work causing spikes and troughs all over the place in the nanosecond range.. crazy. But generally, at _my level_ anyway.. don't care!

If you're running in GHz range, I'm sure theres a lot more worry, but thw hardware is a lot more clever as well; at the MHz range I'm working at, most of the chips you buy are built for GHz speed so not to worry.

I picked up SRAMs that run at 10ns speeds; cheaper a bit (a few pennies) to get the ones that run at 50ns, or 100ns; you can get serial ones as well.. total fail (who wants to access RAM one bit at a time, you want at leat 8 bits parallel!) -- but given the prices were more or less the same, I went for the highest speed; then later someone clued me in that high speed stuff is also subject to other problems.. you'd better keep clamps on the power supply tolerences etc; the higher speed, the higher quality work they expect. Its hardware.. as I say, theres an infinite variety of ways they will screw with you ... but so far, I've had no problems with the 10s parts. If my opcodes take 50ns to run, I don't have to worry about that piece of RAM at 10ns!

Now, every single IC or part has a datasheet, and you end up collecting them all; they're full of ghastly levels of detail (and still miss details you'd want to know).

An SRAM datasheet for instance will have a timing chart ... a nasty picture which shows the variosu ways it is operated, and the timings that occur in terms of data reliability and voltage use and so on.

Consider..

A chip may be Chip Enabled or not; ie: you can turn the CE down and the chip goes to low poweer mode, until you turn CE on again; likewise, you can turn on wrote-mode or read-mode (never both!), so it can know to store a value or produce one; the same data-pins are for both.. it just raises values in read-mode, and latches and stores values in write-mode. So the chart will show what happens when you CE, then enable WR mode, and then raise-data pins; if you do it all 'at once' (say), then how long until chip is awake? how long until its ready to store a value? if it starts storing the value, how long must you hold the value there? how long until it can cycle to the next activity? Just thinking about having to put NOPs or timing delays into the logic so that will work makes my head hurt... except I bought 10ns parts; I can raise CE, and then raise Write-Mode, and then push a register into my data-pins and address-pins.. theres say 4 cycles down, which is 200ns; the chip was ready loooong ago No worries!

But you get into this train where you're worrying about the shape of the crystal clock, or this ir that.. leads to madness. In general, don't sweat it.

In my GPU at high sensitivities, I do worry, but inserting a NO here or there seems to smooth things out, and I need some more capacitors in there.

What buggers me up is .. theres mono caps, electrolytic ones, the ceramics, the dipped tantalum ones, etc; all these kinds, and some are polarized, some are not; each have different cost paramaters, and different Farad amounts; how do you know which you need? Tip #37 -- get the bigger ones, then dont' sweat it If it costs a penny more, its okay to us hobbyists; the big guys buying 10million mono caps are in trouble, but for us.. buy a handful on ebay and dont' look back, dont' worry about polarity if you can avoid it

jeff

 

[skeezix]

There is the invisible cost.. the 'hermit' factor

My wife is sick of packages showing up like Christmas, every other day from far away places; cost is low, and 'free shipping' from China is freaky

 

[skeezix]

Awesome, my AVR Dragon showed up today! (got two shipments today.. big box of parts from China, and a little box of goodness from Digikey in the US. Digikey are pricey, but _amazing_ .. order, ship same day, arrive a day or two later.)

The box on this thing is cute:

http://www.google.ca/url?sa=i&rct=j&q=&esrc=s&source=images&cd=&cad=rja&docid=z9js2hfNlsijBM&tbnid=kcdlGOgSKys25M:&ved=0CAUQjRw&url=http%3A%2F%2Ffab.cba.mit.edu%2Fclasses%2F863.12%2Fpeople%2Fcharles.fracchia%2Fwk6.html&ei=05CbUYi1IK3IywGQ64D4Cg&bvm=bv.46751780,d.aWc&psig=AFQjCNHBl1agTxgnfd19-WgrW8yiV9nIZQ&ust=1369236034948538

The Dragon is the ISD .. In-service-debugger, so you can drop breakpoints and view registers and such live while the circuit is running. I think the Dragon can also do high voltage parallel programming, which would let you 'unbrick' an avr that you've screwed up. ie: The avr's have a concept of 'fuse', which is like a persistent setting.. some of those settings are things like 'don't allow programmers to update you anymore', or 'heres the clock crystal type we're expecting', etc.. screw up those settings and your chip is bricked. You can wire up a cheap programmer, or even use an Arduino or the like, but its a bunch of extra work to make something 'high voltage' so it can unbrick these bad fuse settings. When the ISP (in-service-programmer) and ISD (in service debugger) options from Atmel themselves are 'cheap' (about $40ea), I don't mind just getting the official and well supported kit; you can make your own pretty cheap, but when I was just getting going a few months back I wanted to minimize the surface area of problems.. having my programmer funk everything up would be bad news Anyway... yay, finally got a Dragon!

Since I'm doing chip-to-chip communications between cpu and gpu .. sometimes nice to stop it and see whats going on, especially when I've not built a serial debugger interface yet .. with this, maybe I can get by a bit longer before building that.

The actual pcb itself looks like:

http://www.google.ca/url?sa=i&rct=j&q=&esrc=s&source=images&cd=&cad=rja&docid=9SAlHEsiAlaYSM&tbnid=5NnI7ULcUIM_aM:&ved=0CAUQjRw&url=http%3A%2F%2Fwww.atmel.com%2Ftools%2FAVRDRAGON.aspx&ei=ZJGbUbiyOqXCyAGV3oHACw&bvm=bv.46751780,d.aWc&psig=AFQjCNHBl1agTxgnfd19-WgrW8yiV9nIZQ&ust=1369236034948538

Pics are random web links, not from my actual unit that looks identical.

 

[Asiyura]

Here is my best electronic circuit board (and the only one)  :lol:

I really enjoy reading your posts, it motivates me to get back working on my breadboards!

 

[skeezix]

Nice I've not made a board in 20 years, and that board was etched in a bathtub, with the acid resist drawn by hand. Ghetto as hell.

I'm trying to prototype the hell out of every step since I'm a noobie, and building confidence in the design; once I finalize a schematic in a few months, and send to a boardhouse*, we'll see Pretty complex board for a 'first project'

*boardhouse -- well, maybe I'll photo-etch and print it from a printer, and etch in the 'bathtub' again, so its more real-time; my dad has a spare 'mini drill press', for the drilling..... or maybe I'll send to a boardhouse, if I can find one with a fast turnaround thats not hundreds of dollars..

jeff

 

[LineOf7s]

Just wanted to chime in and say that while I've no idea what anyone's talking about, I'm really digging the enthusiasm being expressed here.


Keep it up!

 

[skeezix]

A note on "CAD" software; its there to help you design your circuit diagram (schematic) and your actual board layout (pcboard).

Usually you can get a combined package that allows for 'schematic capture' (to draw the schematic), that also does assisted pcboard layout. Sometimes they'll toss in 'circuit simulation' which lets you perform some basic prototyping purely in software (amazing!) -- the more advantaged packages are so good in the simulation that you can go Schematic -> Sim -> pcboard and send it to a boardhouse to get a real pcb made, without ever actually needing to assemble a prototype yourself. The enterprise class CADs that cost $30k are so good they simulate all the microprocessors etc right there, simulating the whole board.. but even some of the cheaper costware options can pull off some of those tricks. Anyway, a combined package will let you pick an item from a library and drop it into your schematic page, and that library item will also have a physical pcboard footprint item associated to it. ie: The schematic symbol for a resistor varies by country but is usually a rectangle or a zig-zag; but on the actual pcboard it is a long tube, or an upright tube, or a surface mount tiny box, or.. etc..; likewise, pcb layout will assist you from going from schematic 'connections' to actual routed wiring (which has to go around components or through various layers of the board, unlike the schematic where they can crisscross or whatever.)

To anyone outside of the community, there is a lot of madness here.

- you would think there woudl be some standards for object exchange between applications; I mean, how many hundreds of times to engineers want to redesign the same actual component for each application? Well, turns out these people love to do it.. every app will have its own library set, since they are generally incompatible. (There are some translation tools that are hit or miss.)

- you would think the component companies would publish library files; "here is a full set of texas instrument EagleCAD libraries'.

- you would think library names would be standardized (and standardized component names within libraries)

- you would think, given the above, you'd have central repositories of libraries

Some of this is partially done here or there, but nothing comprehensive. I must assume its greed driven.

Instead.. each CAD tool comes with a limited library pile (though the better ones like Eagle do include a fairly comprehensive set.) Given a part number (read it off the top of an IC say), its _hell_ to find the part in the library; ie: you have to attach one of 200 libraries into your project (each one making your project slower to work with); searching only goes through the libraries you've imported into your project, so when you're new you end up importing every damned library just so you have a chance to find them. Some of the better tools still have a search function that looks at 'first characters' not 'any characters in the name', so if you dont' know their naming standard, you can't find the part. Remember that you might have a Texas Instrumentsd 74LS323 chip, which would be called SN74LS323-Something, but that won't turn up. If it does turn up, its probably the wrong package version.. the surface mount? the DIP? which DIP spacing of pin and body width? what height? .. etc.)  Turns out that 80% of the time, your component will not be in a library... despite the component being a 10 year old highly common one.

Why? Because these apps cannot share. Usually the workflow is for peopel to pull in libraries, copy the components into their _own_ libraries, and work from those; so every single engineer has custom built up their own libraries based on libraries included with their favourite app, or from libraries they've found online. You know what happens then? Dependancy hell -- if those source libraries get updated, no updates flow down to your libraries, so you hope theres never a change, either. (Software guy here, I think like this.)

Well, then, the Design Spark guys are cool, they built a website where yo can try to share a bit; you can upload some components, or download; you can pick a format, to see if they've been translated into your preferred app (the one they sell , etc. Great, cept their search tool _does not work_ (put in some checkbox criterion, and it returns 0 results, even though it shouldn't.) This has gone on for weeks, so its not a temporary glitch; likewise, if you pull down a component from there, it almost certainly will not work, or will be shoddy, or will be the scfhematic side only and not the associated pcboard footprint side, _etc_; theres so many ways for it to go wrong, that they basicly have.

So.. most people seem to design their own schematic symbol and/or pcb symbol; its not hard to do, so they just do it.

Yeah, we want to make our own fonts when we buy a word processor, right? Its like these poor schmucks are stuck in 1983 still.  I mean, this is the day and age where entire operating systems and every conceivable package can be fetched from a repository for automatic install and update into your operating system. Thats because software people are good at hammering this stuff out; sure, you'll end up with dozens of competing standards, but evolutionairy pressure will tend to make a few win. Not so when hardware guys get in there.. they're not good at designing software

Annnyway, so, great, so CAD software is totally what you need, and another learning curve you get hit with hard, up front. Still, learning about 'connections' and 'nets' (connections that are implied, but not necessary to render and clutter your diagram), and then having the software show you the pcboard layout with 'airwires' (unrouted wires) and auto-routing tools and manual routing tools to help you lay the wirign into the boards is awesome; routing is really difficult, and there are again the usual hardware-tricks-you-to-hell things to learn about. For instance, you can' just have a wire trace turn 90 degrees .. the two sides of the trace may start inducing current in each other or other problems; so instead you do turns like the edge of a stop sign .. in steps. Auto-routers can be brilliant, or make a mess. Well and good.

I've tried a few of the open source freeware guys, and despite having strong followings, they're terrible; it seems to exist from a mono culture -- CAD people know how to use CADs, so they don't know or don't see how terrible they are. New people are strictly forced to learn terrible UIs and work with crippling disabilities, since thatys just how it is. Some things are annoying but you can get - what, no resize? you have to show this tiny little item, because fuck you, thats why? oh, its so it fits on the grid bette3r? ....... - but you get it when they're referring to the pcboard side and no resize there, so theres just no resize option at all. (From people who used graphic software like Illustrator or Corel or Inkscape or whatever, this all seems bizarre.)

So..

- Ignore the open source guys; garbage. "Stuff like 'hold down C while clicking an item to clone it'.. yeah, thats standard UI and obvious! Stuff like mouseheel scrolls vertically, rather than zoom in/out like every other application ever made. Etc and so on; every single facet seem to be tedious in these guys; the software _works_ and is featured, but the UI interface design choices are insane; therefore I must conclude its a 'all CADs are like this, so they made theirs suck too'; a lot of people bitch on their first use of these tools, from the forums  etc.. .. how much time do you want to spend tripping over your feat, when you just want to get in and be useful?

- Theres DesignSpark which is _free_ commercialware; you sign up for an account,a nd get the app for free; they encourage you to use their other services as payment,s o all told.. pretty decent of them; the software is not too bad. But see abov for library hell, and a lot of things were pissing me off.. but ultimately, Des Spark is pretty good, out of a world of crappy apps its one of the least-crappy.

- Eagle CAD is the bigger name one; theres big expensive packages, but Eagle is one that is generally well respected, generally standard in that all boardhouses and shops use it or can support it; you can get a free demo with limit to top/bottom layers and a small board (like 4 inch by 3inch, something like that; like an arduino size.) If you want to pay some bucks, liek the price of a game, you'd get a few limits removed; but if you want th eufll package, its oging ot cost like Photoshop.. hundreds and hundreds of dollars. But they give you the full freebie for small boards, which is pretty decent of them.

Eagle has a fairly good default library, though its again hard ot find your bits in it. It tends to make sense to me more .. mousehweel zoom as god intended No, its like PHotoshop or Gimp .. its got all the goods oyu need, very powerful, but a lot of cryptic UI; but most of it tends to _work_ at a basic level,and since its one of the big guys, theres thousands of youtube videos with turoials, suport forums, etc etc.

So I'd highly sugest Eagle, or Design Spark.

Screw the rest, and screw these guys if I could

jeff

Disclaimer: Remember, I don't really know what I'm doing; I'm bashing my head through same as anyone, so what I say is not really worth anything; but its from a bit o experience, I've burnt some time, and I hope I can save you a bit of time.

 

[Asiyura]

Indeed, for people used to Photoshop and other graphical tools, Eagle tends to makes more sense than most of the other CAD tools. It's the only one I could more or less use, and I've tried opensource softwares and even designsparks (that, at the time didn't work well _at all_).

And Eagle is free for small projects (and if you want to make a large project, you can always make a few boards at the max size )

I've got TI's "IDE" for MSP430 installed on my windows virtual machine, ready to review my old code and read the tutorials that I made and/or translated from english to french. Tomorrow will be a mix between learning HaxeFlixel for my game and re-learning to deal with µcontrollers  h34r:

 

[skeezix]

TI is great; they sent off a few thousand mcu's for free to anhyone who asked awhile back; I got one of those, that has a USB carrier.. plug it into your PC, and theres eclipse and gcc etc ready to use, .. program the chip, pop it off the carrier, good to go. Love it

Sorry I just cost you all your free time

 

[levi]

I've got one of these on preorder, which I'm hoping will be a gentle entry into hardware hacking for me.  It's a little debuggable board (so you can debug it with gdb over usb and pretend it's a program running on your computer) with an ARM cortex-M chip on board (I'm a fairly old hand at ARM coding, so I'm planning to try coding directly on the metal there) and once you've got that running you can experiment with driving more than a few blinkenlights off it.

Course, clocked in at 72MHz and with everything built in it's cheating a little bit, but I'm not that interested in using old hardware myself - slow ARM chips are ten a penny these days, and actually looking at it, most cheap configurations come with most of the stuff that this board claims anyway.  Looks like they're about the same price as Z80s, although on the Farnell site I'm unable to find anything in a non-surface mount package, while Z80s can be got in DIL format quite easily.

 

[skeezix]

Like I say, soldering SMT is actually easy.. maybe easier than DIP anyway shockingly - if you've got a board to solder to (you need that solder mask present; then you can just drag-solder across it and the solder will magicly head over to the pins due to surface tension and the solder mask.. magic!) .. but yeah, ARM stuff is nearly all SMT, due ot high volume of pins, and small size. I've seen them in the _hundreds_ of pins.. yike!

You can often get adaptors (break out boards) .. I ordered a dozen from a nearly-local shop (http://www.proto-advantage.com/store/) so they were only a couple bucks to shiop, and a few bucks each adaptor I needed. Making your own adaptors is also doable, but gets you into the CAD/design/print/etch loop right off when you may not want to; so I ordered some SOIC32 adaptors so I could use these 10ns SRAMs I'd ordered, and made life way more easy for prototyping.

Anyway, yeah, looks not too bad; I might well end up with an ARM chip for my GPU, since the speed is certainly nice .. though I've got a couple of these high speed 8bit SX28's around still I can try to use.

Also, fwiw, breadboarding with 70MHz will be trouble in my experience; SX28's use SXkey programmer, and it has a clock generator right in it, so you can use the UI and drag a slider up and down to set the clock in real time; so if I dragged it up past about 30MHz things started to get weird.. hti about 40MHz and it'd crash hard, on the breadboard. Funny though since I've seen some projects working at 60, 70, 80MHz on BBs, but no dice for me. Perhaps it depends on other factors or I'd goofed somethign, but I had no luck at these high speeds. wikipedia and some other articles I've read suggest BB's max out around 10MHz or so, even, though I've had no problem at 20MHz.

I just picked up some 30MHz crystals, but my avr8s are rate donly to 20MHz, but sticking in a crystal to overclock can't hurt .. so maybe I should see if they work; would make my gpu a lot easier to work with But if it doesn't work, wouldn't know if it was due to avr8 limitations, or BB interference.

Anyway, so yeah, hope that sucker works out fine in the BB for you, or maybe you'll have to underclock it if you can for that environment.. something to ask the KS guys!

jeff

 

[skeezix]

Found this with some googling:

http://ca.mouser.com/Search/Refine.aspx?Keyword=LPC1114FN28%2f102%2c12&FS=True

Thats a low end ARM chip with DIP.

But like I say.. on ebay I found some guys in Thailand (we'll see when they arrive TQFP adaptors all the way up to TQFP100 (!) .. thats two rows of 50pins, so not be fun to wire into a breaadboard, but certainly doable. (You also likely don't need all those pins .. most of them usually are left for user function, such as address control lines and such; if you're not addressing meganbytes of RAM or driving a dozen PWM motors, you'll likely need only a couple dozen lines anyway.)

jeff

 

[levi]

Trouble is, the only things I've been able to find on Farnell are in various flavours of BGA which I've never even considered soldering by hand before. Designed for reflow I assume. Everything in plain forms of SMT are out of stock. Would be interesting to see the pinout on that DIP chip you found given it's only got 28 pins.

I'm hoping that prototyping with these things once you get them soldered to something shouldn't be too tricky, as they got so many connection options you don't necessarily need to use the full speed clock externally. If you can get away with hooking things up using i2c on that one they use on that Galago board for example, I think the clock runs at only 1.5MHz. It's only if you want to attach extra RAM to supplement the on-board 4k that you'll need a faster clock, but I've less than no idea how to go about doing that, so I suspect I'll be happy with the on-board RAM and slower interfaces for a while, especially as given it's got more flash ROM than RAM I assume it's using NOR flash, so I can run my program directly from the flash.

 

[skeezix]

Generally chips that have flash built in, can run from it; theres a trick there.. conventional systems have a single flat space, but these guys usually have flash, plus local store; maybe even external RAM too.

Consider:

char *foo = "bollocks";

Is 'foo' in flash, or in your heap/stack? Sometimes you have to care

avr-gcc has some special annotations you can make to help identify locations and so on, so I bet you'll have to dig into things a bit.

As to adding RAM .. well, I've found out it is not too hard at all to wire up! Very easy in fact! I eluded to it above, but I can write up a short article about it. You can buy 32K or 128K chips cheap.. a couple of bucks each; you can get MB chips and so forth, but you probably don't need it.. not much more expensive than K-chips of course.

The trick I found is .. tie high or tie low the unused address lines, or they _will_ fluctuate randomly and svrew up all your transactions; if you need only 'say' 8bits of address line, and have 16 or 17, tie down the rest; then you just need to toggle Write, or Read, or Chip Enable, and they work as expected.

Interestingly, Flash chips have the same pinout(!), so if you like, you can drop in SRAM or Flash, same.

(EEPROMs and such are different pinouts of course, since they use high voltage for clearing etc.)

So yeah, adding SRAM.. you'll be fine I just hit up digikey and add some filters.. package type, speed, size, and narrow in quick; I can give you the part numbers of the ones I use but theres a billion SRAMs, and they're all good.

jeff

 

[TrashyMG]

Trouble is, the only things I've been able to find on Farnell are in various flavours of BGA which I've never even considered soldering by hand before. Designed for reflow I assume. Everything in plain forms of SMT are out of stock. Would be interesting to see the pinout on that DIP chip you found given it's only got 28 pins.


I'm hoping that prototyping with these things once you get them soldered to something shouldn't be too tricky, as they got so many connection options you don't necessarily need to use the full speed clock externally. If you can get away with hooking things up using i2c on that one they use on that Galago board for example, I think the clock runs at only 1.5MHz. It's only if you want to attach extra RAM to supplement the on-board 4k that you'll need a faster clock, but I've less than no idea how to go about doing that, so I suspect I'll be happy with the on-board RAM and slower interfaces for a while, especially as given it's got more flash ROM than RAM I assume it's using NOR flash, so I can run my program directly from the flash.

http://www.mouser.com/ds/2/302/LPC111X-184855.pdf

Page 23-25

 

[levi]

http://www.mouser.co...111X-184855.pdf Page 23-25

Cheers! So, if I'm understanding it right, so long as you poke the right value at the IOCONFIG address, you've got 12 GPIOs, or a bunch of ADCs and I2C and other bits and bobs.
Also from the address map on page 43, flash is simply mapped as starting at zero page, but readable as usual memory. Dunno what happens if you try to write to it of course - hopefully it just interrupts you until it's done its business, or transparently buffers the writes, but knowing flash it's probably messy and requires your program to handle the block erase and rewrite stuff.

 

[WizardStan]

Attempting to write to flash usually crashes, like a segfault. At least that's how all the microcontrollers I've used in the past handled it.

 

[levi]

Ah right, that's fair enough I guess. It's not like it's secondary storage, but just being used to simulate a masked ROM or EPROM.

Anyway, I should probably stop spamming up Jeff's thread with this talk about ARM pinouts. Go go super zikzak!