Portable diagnostic link

robm

Billy Mays
Staff member
I made a brief mention of this over at BH in passing, but now that I'm nearing the end of my N64 project (finaleeeee!) I'm getting ready to lay out plans for the next step in my N64 portable plans. Beforehand, I'd like to get a feel for what kind of demand this would have - I'm going to do it mainly as an exercise in microcontroller experience, but if there's people out there who wouldn't mind having the same benefit for their portables, then I may make it a more universal set up instead of just for the N64.

Here are my plans, thus far:

Code:
N64 side:
   Connectivity through the LOB64 port to cut down on holes required to cut
           This should work just fine with an installed switch somewhere to switch betweeb LOB64 and diagnostic function.
           May be forced to go with a mini-USB port instead, which wouldn't be that big of an intrusion.
   Designing required circuit to sample processor temp, 3.3v and 12v line measurement, and digital signal detection to 
        determine if the N64 processor is, in fact, "talking" to the cart.
   Current processor I'm aiming for is the Atmega128; it's small, powerful (likely much too powerful for what I need it for, 
        but eh,) and doesn't take much power at all. Break-out boards are cheap and easy to use. 
   I've begun minor development with the Atmega168, but the profile is a bit too large for my taste, as portables these
        days need all the room they can get! Code migration between the devices isn't difficult, so will likely finish 
        development with the 168, as they are easier to deal with if they break.
   
PC side:
   Utilitarian GUI, nothing fancy, just functional
            Have begun work on this in Foxpro, not sure if I'm going to port over to Visual C++ yet.
   USB connectivity           
   Firmware update support (in case of future upgrades, for whatever reason)

It's a large undertaking, particularly as I'm rather early in my microcontroller experience, but I'm pretty excited. The big obstacle that I see so far is designing the N64 package, and its required wiring, to be small enough to fit in the smaller portables that people are making these days. Hence my decision to go with the '128, as it'll proivde a very very thin profile for the application.

My end result in mind is to provide this modding community with the means to have some diagnostic ability for their portables, even if the data supplied is more superficial than helpful. I would love to be able to hook my N64 up to the computer to see how the TI regulator was doing, or to make sure that the 3.3v line is getting proper voltage, without having to deal with opening the case and finding that Dang elusive multimeter. (Ok, so maybe I'm the only one here that has kids that run off with multimeters.) I plan on providing a lot of the service, as I don't expect everyone here is able to flash their own micros. I think providing a supplemental parts list that can be somewhat easily obtained on Digikey, while providing the micros myself, would be a suitable arrangement for those few who might want to tackle this.

Thoughts?
 
What would be the final cost for this? If it is under $10 or so, I can definitely see this being an absurdly cool addition to any portable. :mrgreen:
 
PalmerTech said:
What would be the final cost for this? If it is under $10 or so, I can definitely see this being an absurdly cool addition to any portable. :mrgreen:
It easily could be, depending on how he actually intends to interface with the computer.
Even USB could be done with relative ease.
 
this could be really awesome! if something breaks, plug it in and then it will tell you. unfortunantly, it probably will cost a bit more than 10$ :rolleyes:
 
samjc3 said:
this could be really awesome! if something breaks, plug it in and then it will tell you. unfortunantly, it probably will cost a bit more than 10$ :rolleyes:
And why would that be?

Judging by his description, this would only need a single Microcontroller, and *maybe* a few discrete components.
This could EASILY be <$10
 
jleemero said:
samjc3 said:
this could be really awesome! if something breaks, plug it in and then it will tell you. unfortunantly, it probably will cost a bit more than 10$ :rolleyes:
And why would that be?

Judging by his description, this would only need a single Microcontroller, and *maybe* a few discrete components.
This could EASILY be <$10
hmmm. i didnt think microcontrollers were so cheap. but thats because my dad is always programming custom ones that cost a ton. :wtf:
 
Probably is something well beyond a Microcontroller if it costs any incredible amount of money.

The ones we'd use for a project like this?
< $5.
 
jleemero said:
Probably is something well beyond a Microcontroller if it costs any incredible amount of money.

The ones we'd use for a project like this?
< $5.
well. they are micro, but they are very powerful for thier size, and stuff :tophat:
 
This'd be a great addition to a portable, especially if you know you're prone to messing things up (SHOCKSLAYER). If the portable doesn't turn on or work properly, hook it up to your computer and figure out the problem.
 
Total parts would be more than 10 dollars or so, but not prohibitively so.

Here's the breakout board: ($4.95)
http://cgi.ebay.com/ATMEL-ATMEGA128-PCB ... .m20.l1116

Here's the processor: ($7.00)
http://cgi.ebay.com/ATMEL-ATmega128-16A ... 1|294%3A50

Remaining parts should be negligible; it depends on whether or not the '128 can, itself, sample voltages, or if a circuit needs designed for that. I'm pretty sure the '128 can handle it, however. The remainder of the cost would be wrapped up in the temperature sensors for the processors, which again shouldn't be expensive; interfacing those to the '128 would be another interesting experiment.

I would say, for the end user ordering the board and processor from me pre-flashed and ordering the remaining parts themselves, it should be well under $20 max. The board for the microprocessor would already be populated with the required components. The PC front end would be provided as a free download for all, of course, and would need little to no customization itself. My end intention for the package would be a "Connect this wire to desired voltage A, this wire to desired voltage B, this wire to desired voltage C, and this wire to desired temperature sample A and this wire to desired temperature sample B."

That still leaves a whole lot left that the processor could do, but eh... "future expansion," as they say.
 
Woah, why use that QFP + Breakout board?
The mega128 comes in a DIP package you know, much smaller than that breakout board, and not to mention cheaper at only $7 or so.
 
The dip package would be a whole lot taller, which is what I'm trying to avoid. The '128 is extremely small, so the size of that breakout board SEEMS like it's monstrous, but in reality is less than 2" on each side. Total height around 1/4", maybe.

After some additional research, I smacked myself in the head. The AVR does have eight ADC lines, so sampling voltages will not be a problem.

The good news about the AVR class of microcontrollers is that code migration isn't difficult, so on down the line those who would like to use simpler devices may do so easily. The Atmega168, for instance, is only $1.95, though I'm not sure if that particular device has ADC. Toss in the rest of the components required, should be will under 5 bucks, if you've got the parts laying around your house.
 
Uh.
DIP == 3mm thick.
Obviously with the leads it's a lot taller.
But all you do is bend the leads out :p
It's what I do for most any modchip install or anything like that!

But regardless, either way is fine.

And besides, if you do your job right when programming the thing, then it should work with either package anyway ;)
So I've nothing to worry about :awesome:
(You're clearly knowledgeable enough, so I don't expect you to do anything "stupid" that would mean work for me in the end)
 
Ah, yes. MmHmm. I totally understand this.

But you should make an easy guide when you are done, so that, you know, dumb people can understand. Not me, of course. :p
 
:lol:

Do not worry, full documentation will follow.

The dip package just isn't my choice of design these days... I like SMT stuff, looks cleaner and more professional.
 
Ok, if it's cool I'm going to go ahead and start a WIP in the "Other" section, as the application will work for any portable system.
 
robm said:
:lol:

Do not worry, full documentation will follow.

The dip package just isn't my choice of design these days... I like SMT stuff, looks cleaner and more professional.

I hate SMT. Impossible to solder. Oh wair, I just suck.

Anyway, I've been thinking of getting some microcontrollers to play with. But I don't have a lot of monies. I would need a programmer, microcontroller(s, I'll break them) and maybe a development board. Any ideas?
 
To be honest, a dev board is nice to have but not necessary at all. You can get an AVR micro and an ISP (in-system programmer) for around $30, depending on where you got the programmer. If you don't already have a breadboard, you can get one of those from a local Radioshack outlet for relatively cheap. I think mine cost me $11, but it's been a long time.

Beware of knock-offs from eBay. If you can get a USB-based programmer, that would be ideal, as a lot of people end up with problems with the serial-based programmers for some reason.

Here's a programmer that would do fine:
http://cgi.ebay.com/AVRISP-Atmel-STK500 ... 1|294%3A50

Here's a chip that would be good for development on a breadboard:
http://search.digikey.com/scripts/DkSea ... 68-20PU-ND

For some reason, I can't remember where I was getting mine for $1.95. If you've got the money, you could just buy a kit from Radioshack for about $70 - $90, but I don't feel you learn as much with those as opposed to doing everything yourself.

SMT used to scare me, until I got good using a toaster oven. :lol: My main problem is trying to keep my hands steady, though that's not really as much of a problem, because when the solder melts the parts "align" themselves to where they're supposed to be. Pretty cool stuff.
 
robm said:
SMT used to scare me, until I got good using a toaster oven. :lol: My main problem is trying to keep my hands steady, though that's not really as much of a problem, because when the solder melts the parts "align" themselves to where they're supposed to be. Pretty cool stuff.
Nice.

I still do all my SMT work by hand. :tophat:







:cry:
 
Seriously man, go get a toaster oven. They're 20 bucks new, if you get a two-layer one. They're even cheaper from places like New Uses.

There are tons of tutorials on the web on how to use them; there are self-heated skillets you can use too, but toaster ovens are more uniform.

I don't even think I would attempt a hand-job these days.

...that's what SHE said...
 
Back
Top