My final year project...

[bsodmike]

Hiya guys,

I guess the geekier few here might enjoy this, or completely think it is pointless.

Basically I have implemented a working PS/2 'host' protocol on a microprocessor, and the reason for doing this is so that we can use current 'optical mouse' tehnology to act as a Tachometer replacement. Of course there is a fair bit of optics involved so as to adjust the focal length, DOF, and FOV etc and I only start testing on Monday



For more pics and a bit more info check out http://www.bsodmike.com where it is my main post right now.

Cheers,
Mike

[Jonny M]

I don't want to crap on your thread here Mike, but can you please post the full article? Nobody minds you linking to your site in your profile and sig, but using the forums to send traffic to your AdSense program isn't on.

[bsodmike]

Oh sorry Caged, I thought it would be silly to repeat it (and argh, got to relink the pics heh)
--------------------------
In September of 2005, Dr. Trayner contacted me as he wanted me to ‘use’ current optical mouse technology and attempt to turn it into a sensor for his MEng project RABOTA - a search and rescue robot. Little did I realise how lucky I was…

Even though this is my final (third) year project I have a huge ‘team’ work factor in my project as I collaborate with 5 other MEng students who are developing other subsystems of the robot such as internal communications, driving the motor, battery management, lighting and the ‘host’ control computer. The idea is that my sensor will communicate with the host and send it data from which the host will be able to figure our how far down the tunnel the robot is.

And now for some images

This is a shot I took last week of my setup:



Quite sadly I wasn’t having much luck and it simply refused to play dead…



To give you an idea of what is wrong here, basically, when I issue a ‘0xFF’ which means reset, the mouse should respond with 0xFA (acknowledge), then 0xAA (self test OK) and finally 0×00 which tells me that I’ve connected a mouse and not a keyboard (0×01).

But today, after implementing ICSP, I got a break.. After a bit of work, I was able to tidy up the output that I was getting via serial coms:




The LCD was implemented last semester (with a some other stuff) to allow simple debugging while working on the robot…



This is my little niche in the MEng lab



I present to you RABOTA, er…well….it’s all we have of it right now..I think hehe.



Yesterday Dr. Trayner told me that he would have the 1/8th scale model for me to test my sensor with ready by Monday and my code wasn’t done and I confidentally told him “I’ll have it ready by Monday” but no way did I expect to have it FINISHED by tonight!

I’m simply over the moon and really am quite happy that my code works…ahh the feeling of getting something done right is pretty sweet. I have also sent the final PCB design for prototyping and I should receive the final PCBs by Wednesday next week!

[Steve]

What, no oscilloscope?

SHAME ON YOU!

[Jonny M]

Cheers Mike, the more original content we can get posted on here the better. It's also helpful for those times someone might hit up your HEXUS post through Google, if this is a few months down the line then your main article will obviously have changed and they wouldn't be able to find it.

[Zak33]

good point Caged...I didnt think of that.

Nice work Mr Bsod

[herulach]

Looks cool, i knew i should have done engineering instead of physics, my final year project is the bloody wigner representation of the QM density function, or 3 100 page books full of integration as its more commonly known

[smtkr]

What kind of ucontroller are you using there? It seems like I have done some work with a PIC and LCD that look just like that years ago (I really hate the register set in those things).

[bsodmike]

Lol STEVE - I used the osc scope last semester but I knew I could get it working with out bothering with it and didn't take to long really. The clue to this was that I was getting 0xAA and no 0xFA, i.e. I was sampling data a full byte too 'late'...

I'm now gonna work on a python prog to calculate the maximum velocity reached by the mouse - actually I'm doing this in 2 parts:

1) Attempt use the TMR0 interrupt and have the PIC work out the max velocity

2) Send the raw packets over RS232 and do it in python (easy).

...I should have #2 working by Monday Oh I'm also taking the resultant vector of the x/y packets i.e. pythagoras sqrt(x^2+y^2) - my code went from 1500 words to 2700ish with math.h, divides and 'sqrt' haha...

I'm using a PIC running @ 20Mhz and the LCD is driven in 4 bit mode (control it via two nibbles).

Cheers, Mike

[Agent]

Thats quality Mike, keep it up

[bsodmike]

I seem to come up with the craziest of ideas with less than two weeks before I go home for Easter :/

Just desiged a little PCB that'll stick onto my main device via the DB9 connector (just slip in really) and it'll log the data sent from my device to an EEPROM IC via IIC. By default it'll be in 'read to store mode' but if it is removed and connected to a PC I can send control commands and have it a) play back all the data that was sent to it, or b) erase EEPROM

I think this is going into my report under "Future directions"... heh.

[bsodmike]

I’ve pretty much spent around 9 straight hours in my ‘resurrected’ makeshift lab that is currently confined to a long bench in a fairly dark room which has two antique beds as well. One of them serves as my second bench haha.

The final OMO PCB arrived yesterday and this had a few revisions I really didn’t want to leave out in the final job as it is a much much more professional PCB than the rest. I’ve fully soldered it up and even though I was initially lacking a few parts I eventually found what I needed while rummaging though some jiffy bags from Farnell, including PCB mount switches.

The progress was a bit slow as I had to crimp the cable for the OMO to LCD which has 32 pins and the ordering had to be spot on. I also decided to solder up the D15 connector to allow this to be plugged into the robot straight away. Used plenty of heatshrink and even threw on a sleeving of cable braid for that extra ‘professional’ look and it actually feels great.

I needed a source of power, +5V to be pedantic and I found a random AC transformer and opened it up. To my shock and horror their idea of a ‘power supply’ was 1x transformer (probably 20:1 ratio), a 3300uf cap and 4x power diodes (1.5A rated)….wow. Soldered on a LM7805 for a regulated +5VDC output and it fried…as I applied -12v to it by mistake, and I could see the output go from -5v to -2 and *crackle* the lovely fragrance of burning plastic…re-did it the right way around after actually testing which wire was the positive from the PSU (rather than following ‘regular’ markings) and bam the OMO works!

I feel that quote with “and I present to you, the fruit of my loins…” is kinda appropriate here:







Note: The text is jumping to the next line after the 8th character as the PIC has my Gold Master code which was designed with a 1×16 LCD that had no support chip, which meant that after the 8th character the rest of the characters are treated like they are on a second line, so it is initialised as a two line LCD…

I’m feeling a little woozy since I’ve pretty much crimped approx 32 itty bitty crimp ends, resoldered a LM7805 twice as I fried one by mistake (hah I really need more light in this room). Rather happy right now and am currently processing the images on my phone and need to take a few more shots for the report. Tomorrow, I’m off to buy a DB9 ’socket’ if possible and then solder up the EEPROM logger. Then to DHL in the evening to get this off to my professor in UK

[Odc]

You lost me on "Hiya"
lol

[monkeyville]

Thats awesome dude. I assume thats MPLab your using . What do you code in?

I'm currently working on a Microchip PIC18F452 in combination with touchscreen panel and a few sensors to produce a realtime weatherstation. Been a nightmare project so far.

[Kumagoro]

I would say dont go home for easter there is never enough time to do what you want
and 0.2% could be the difference between your grades. It is likely that you wont get
to do something quite like it again so enjoy it while you can.

[bsodmike]

You lost me on "Hiya"

Thats awesome dude. I assume thats MPLab your using . What do you code in?

I'm currently working on a Microchip PIC18F452 in combination with touchscreen panel and a few sensors to produce a realtime weatherstation. Been a nightmare project so far.

Hiya monkeyville, some how your nick reminds me of monkey island hrm... In those previous images, MPlab is used, but really I could do it in notepad and run the PICC compiler from the command line...you really can't simulate communication protocols with MPLab heh...

I primarily use C but swap to ASM if speed/timing is critical or if the task is simplistic. As there is quite a bit of protocols to deal with (PS/2, UART, I2C) I decided to go with C and it was a wise decision...

Ooh the 18Fs, and wow that sounds well cool - if you like add me to your MSN (bsodmike@hotmail.com only for MSN) we can chat about this I've been thinking of playing with some of the 18Fs for their USB module...but for now I think I'll be sticking to the FTDI USB<->serial chips heh.

I would say dont go home for easter there is never enough time to do what you want...

Actually, I *had* to come home to setup my lab and I've gotten all this done in 9 hours...at Uni it would have taken me easily over a couple days. I've tried my best to keep distractions a minimum but I think I'm doing alright. Now I just have to finish the report, get cracking on my other 2x modules, sort out the Poster for the VIVA and start on the display stuff for the Snake Game...

Thanks for the posts guys...