Dangerous Prototypes

I build in December, a flash destroyer, everything was fine and fun !

But now, 'cause i'm doing nothing in my life ;o)  i'm planning to hack it a bit ...

so my first goal, will be to replace the EEPROM with a clock, same has someone already did and play a bit around...
so basically cutting few traces and  maybe jumping the unused pic pin to the interrupt pin of the clock ...

that will be my first step....

second goal, it might be possible to free few pins of the pic by inserting a multiplexer, and in this way having more hacking fun ...

hints are welcome :oP

I would have to look at the schematic to check, but if the SOSC pins can be freed, you could add a 32.768khz crystal and do the clock inside the PIC. I'm not aware of any DIP-8 clock chips, but there could certainly be one out there.

@ian,
The most popular RTC chip is Dallas semiconductor's DS1307 which comes in DIP-8 package and with I2C interface.
the pinout of DS1307 is same as the socket/routing in the flash destroyer PCB. You just need to cut pin 1,2,3,7 from the routing and add a 32.768Khz in pin 1&2, Pin7-N/C and pin3 to BIOS battery 3V.

sounds perfect!

Hopefully you will give us some pictures eventually of this! :P

first  an exploration session :-)

i already had draft how the flash destroyer works when i build it so yesterday
was the first time i look into the code and solder few wires  to collect some informations.
( i must say my new Hakko soldering station is amazing !!!) 

@ian , @arupbsk
for the clock i got few Dallas (Maxim) DS1337 8DIP before xmas  i needed them for a future hacking project !,
i don't know exactly the difference between the DS1337 and DS1307 tho ...
but arupbsk's basic layout is right,  for cutting traces ...

on the schematic of the FD,  the pin 17 RC6 is free so i might route it to the interrupt of the clock  or 
i might use it for something else (button maybe) ...

so yesterday i look more in deep into how the refresh of the leds are done, well it is done how i already draft.
so placing each segment value and driving each transistor one after the other,  at 209.46Hz (from my logic sniffer)
so each 7segment at 29.92Hz no flicking.

so i moved into looking into my second goal a bit
can i put a multiplexer (3to8)  for switching all 7segments in sequence, and free 4 PIC's pin ...

there where i need more exploration...
 
i suppose a PIC behave like and HCT ic,  so i will have to find out how each transistor are turn on ( i guest in saturation mode)
by the PIC pin, and see if i could turn them on with the output of a multiplexer ....

well we will see next time :o)

specs for  BC 547 are kindof unprecise ...
i think i will have to trace my own ib ic vce etc ...
or at last few useful points :oP

pic of my first session of hacking,  minimal damage was done on that poor FD !

so the most  left led display is now selected throught a multiplexer 74hc138, and
inverter.  the mux addressing is not integrated yet.
the goal was to look if the output of HC ic can driver correctly the transistor and the led.
everything seem fine :o)

i also took time to play with the BC547 transistor, i didn't trace caracteristic ib,ic, vce since a long long time ...
it was fun to do it :-) that little guy get pretty hot when ic is >80mA i stop at 85mA, it sure make sense that the continuous max  is 100mA...

i just this afternoon learn that there is another species of multiplexer ! 
cool an 74xx238  no need for inverter at the output :o) ... but nobody around here has
that in HC. 

mystery question:
a store here has this one  MD 74SC238 AC
ok it is from Mitel, but i can find no doc !  AC is for ceramic package...

so  what is  this  SC family  ????  is it a kind of HC ?  i can find NO spec on that either .... 
any hints from some pro ? :o)



so what next,
i 'll try to fully integrated the mux ...
(alot of trace cuting and board grinding ! :-))
:o)

Catalog: http://www.addison-electronique.com/pdf ... UCTEUR.pdf (http://www.addison-electronique.com/pdf/catalog_pdf/070000-G-SEMI_CONDUCTEUR.pdf)
Cat No: 073353
Description: 74SC238 is nothing but a 74S00 series IC.

lol  arupbsk !

how do you know Addison in Montreal ? did u just google it  ?  lol  :op
this is where i went to buy 4 74SC238 this afternoon !

i was also thinking S for Schottky ...  and  SC maybe  Schottky  CMOS (seem kindof weird) but i didn't find any specifications ...
well  i think i will make few tests on my SC interface...

well, i finally hack my way into the first phase of my planned FD megahack !
right now i just finished the 7 segments hack ...
i have now few hours of software rewrite ahead :-)
(hopefully i didn't make any mistake in the process)

what did you do!!!1 I dont understand. there is a PIC already there and you already know the layout works fine to run the 7segs. So to me the best option (best quickest/easiest) is to simply re-program the pic... Im not pic expert but im sure i could replace it with an AVR do to the same.

I dont understand; i mean if the goal is to do somthing more extreme then just make it a clock then sweet deal! but otherwise im confused by all the wires and minipulation going on here.

It is an amazing wiring job. To me it looks like the perf board is being used to re-route some of the traces.

lol  here the other side :o)



display hack works great, it shows thing perfectly,  need to incorporate a font tho :oP
http://http://en.wikipedia.org/wiki/Seven-segment_display_character_representations
 
the  ..00..  little guy is just for fun ( i also tested all segments ) :-)
no animation yet ;oP

@BrentBXR
in my first message i had 2 goals,
me second goal was to free some pic pins , to have fun with ...
the first thought was to insert a multiplexer for driving the 7  7segment transistors and create minimal impact
so freeing  7 pic pins(original)  minus 3pins for the mux would had give me 4 free pins...

i was in the mood of over hacking so instead i choose to use 2 shift registers (74hct299)  one for driving
the transistors, and the others one for lighting each led of a segment , with big mess of wires ..
so 15 pic pin (originally)  minus 4 pins (for driving my 2 shift registers) i got now 11 free pic pins
for doing ... i don't know what yet :-)

 i could have choose to used maybe an i2c shift register but .. no, i prefer to let the i2c bus free
for playing with things later :-)

@ian
i bought recently an HAKKO FX-888 soldering station,  i'm completely blasted !  i used to used cheap all in one solder pen, there's is no comparison the soldering station is so much more good in all aspects so soldering traces is  fun :-)


so what next ...
of course the main goal, not expressed in my first message  is  " to have fun " until now it is a success on that too i don't think it gonna stop :oP
...
so routing those free pins to header,
writing code for displaying numbers ...
installing my real time clock...

and i already have an application to try it on but it is a secret for now :oP

:o)

i took the time to update my Open Bench Logic Sniffer with Jawi new client and for test i took a peek at my video refresh :o)

top we see pic interrupt inside witch we change digit to display
so 733hz  or each 1.366ms

bottom we see each pulse i make for loading  both shift registers with the appropriate values
time past in the interruption,  around 23us

so there is "plenty" of time for the pic to do some useful stuff :o)

here  those 2 74hch299 and the basic idea :-)


and finally  i soldered the last header :oP
i did not solder the 4 pins of the 299s directly to the pic ports, i will use jumpers, it more Dangerous like that !  :oP

right top corner,  socket for the clock
left down headers for a bunch of free cpu's pins
right down,  28 pins socket and headers for  ... ( currently a top secret project )  :oP


now i got zillions  lines of code to write !  ;o)

i took time to work on my stuff last weekend :-)
so here the first all working thing !

channel 0
high priority ISR for display output disable , while high we change to another 7 segment for display
channel1
clock for shift register299 segment loading ... see previous messages .

i2c on channel2 SDA channel3 SCL
currently i2c bus has 2 chips,  a RTC DS1307 and 16bits io expander mcp23017  (the top secret 28 pins socket was for this guy :-))
so each to those chips trow low priority interrupts  to the pic,  the RTC output an 1Hz square wave and the io a pulse on change of any io input pins.

so channel 6 is push button connected on one io pin, 
the io chips responds on channel4 by a interruption pulse , a little later my low priority ISR talk to the io chip on the i2c bus
getting the new io pin values.
while in this transfert an high priority occur (segment refresh) not problem for the i2c bus ...

on channel5  is the square wave of the RTC
a little time after io activity,  the RTC create a low ISR, then ISR talk to the RTC  on the i2c bus getting RTC values ...
again during i2c transfert there is a display high ISR who delayed a bit the i2c message.

so that is the state of my hacked flash destroyer for now :o)
i got a strange clock with a full mess of wires :o)

So ! ....
2 years later, here an update on living with my dangerous Flash Destroyer hack !!!

First off,  front and back picture :o)





I added few things since its birth ...

A battery for the real time clock,  really handy when you reprogram code and you don't want to lose the current time ...
4 buttons who are routed to the bottom left header, and who connect to the IO chip with green jumpers.
Some kind of feet, printed on my Prusa2 clone :o)

So what is it life since 2 years ??
well this is my principal clock on my desk and I use it from time to time has a prototype board.
So it turn out pretty handy and I still love the look, (and the weird look of people looking at it)  :o)

I added software to helping me with things.
There's 3 different kind of clocks, 
a standard clock, date display, 
another alternating hour - date clock,
and clock with strange font...

After than I have menus to display the read out off the IO chip.
and few entry points to custom piece of code I want to used or test,
related to breadboard use or piece off code I want to try.

Ex.  I used it to test a generic soft timer package, trigger at day-time like... 
I used the IO for triggering a 433mhz  kind of beacon for a while ...
and used input mode to validate a 4 lines status of a thing I built...


So everything was fine until just before last Xmas ....  It suddently started to  reset itself ....
Geez I thought it was dying !  so I tested all chips , just to find the IO chip was going mad after few i2c messages , wierd...
After swapping the chip, it was like new :o)

Since then I used it few days, for adapting the microchip USB stack,  it took me a bit, cause MC, did not fully port their usb cdc exemples niether their stack to the xc8 compiler ... ( I cannot used the DP usb stack for the next project :o( )

So now it is working, the next step is to find an Android driver who emulated a serial port (usb CDC) and link my cheap android tablet to my cool hack !!  Anyone have done that here ? 

:-)

so what is the life of your dangerous things ! ?  :o)