you are correct. But I cant see any other way. If I am encoding my signal in any kind - I will not be able to achive instant response.
But that is a problem that can be avoided by a second signal like a normal "BlueTooth" that the reciever will recieve a signal in the next few ms. So it can be determinded wheter the reciever gets junk or the actual timing singal.

yes - something like that would be good.
It is not needed to feed a wave to the TX. It can be enough to simply turn on the transmitter and recognize that there is a signal on the reciever. Nothing more complex would solve my problem.
But I cant find a circuit that is reliable enough. and all these mainstream modules are digital and have a discrete output/input.

A arduino will be used for further processing. I have other methods to measure the time

thanks for your reply crossroads
Your idea would not match my needs. A 2kHz signal would lead to discrete readings as well. There can be a timegap from 0.5ms (1s/2000) to 0.0ms which is unknown.
In my needs the timegap should be always constant, because I am measurening something in the range of sub nanoseconds.
Thats why I need something continous/analog. Once I know the time it needs for the transmitter to send and for the reciever to recieve the signal, I can use that information for my calculations, that rely on really small amount of time.

Hello

For my present project, I need a RF transmitter and a matching reciever, to trigger something very time important. I dont need to transmit any data. Simply a signal that is there or not there, or a loginc "1" or "0" every second or so. The problemm is, that the signal must be strong enough to pass easily through some solid walls, and it must be countinous. I dont want any discrete logic refreshing with 10Khz or something like that. I want a analog circuit, that responds continuesly to the transmitter. So If I compare the input "1" with the responding output "1" of my reciever on an oscilloscope, there must always be the same time difference. Apart from that the transmitter and the reiciever must work very responsive. (total time of signal transmission should not exceed 1ms) Do you know any IC or circuit which would match my needs? (I am thinking in 300-1000Mhz range)
Instead of a normal RF - Signal with fixed frequency, I was also thinking about an elctromagnetic pulse, which is powerfull enough to be sensed through a couple of walls. - but I think I might get problems woth other elctronical devices near by.
I have already tested mainstream ASK 433Mhz rf modules under my oscilloscope, and as expected the time difference alternates, because of their discrete digital sensing.

Thanks in advance!

regards Max

no one of these want to work for me...
compiler says:

Code:

error: 'class File' has no member named 'readBytes'

I am using the SD.h library with arduino r23

here is my code:

Code:

void f_SdToArray() {
  sdx = 0;
  sdy = 0;
  myFile = SD.open("data2.bin");  //open file 1
  if (myFile) {  //does if file exists
    while (myFile.available()) {   // reads from the file until there's nothing else in it
     // FRAME_1[sdx][sdy] = word(myFile.read(),myFile.read());
     myFile.readBytes(FRAME_1, 4096);
      sdy++;
      if (sdy == 256) {
        sdx++;
        sdy = 0;
      }
      if (sdx == 8) {
        sdx = 0;
      }
    }
  }
   myFile.close();    // close the file
}

hey techylah! thx for your reply!
All your discribed methods give an compiler error...
readBytes() is not understood by r23 nor 1.0
myFile.stream.read() is also not acceptet by my compiler
only myFile.read() works.

Here is my current SD function:

Code:

void f_SdToArray() {
  sdx = 0;
  sdy = 0;
  myFile = SD.open("data2.bin");  //open file 1
  if (myFile) {  //does if file exists
    while (myFile.available()) {   // reads from the file until there's nothing else in it
     
      FRAME_1[sdx][sdy] = word(myFile.read(),myFile.read());
      sdy++;
      if (sdy == 256) {
        sdx++;
        sdy = 0;
      }
      if (sdx == 8) {
        sdx = 0;
      }
    }
  }
   myFile.close();    // close the file

data2.bin is exacly 4096 bytes bif and my array (2D) FRAME_1 is [8] [256] (4096 bytes as well)
It would be nice to achieve it whithout that massive counting´n loops and that stuff.

Anyway - I have still no clue why some interrupts interfer with the SD library... Are there some SPI interrupt flags set, which can be used normally as well?

grz mx

something really strange I found out now:

I am using the SD.h library on Arduino IDE r23.
I am using some interrupts and timers for my sketch.
The problem is, that the SD library crashes when the timer 1 overflow intrrupt is activated
-When i deactivate it it works very well (atmega 1284p).
Has anyone a idea??

My sketch now reads an .BIN file (i filled it up in an HEX editor - random values)
the file is 4096bytes big, and it is transfering at a speed of around 20kByte/sek.
- That still seems a little slow to me - I need to investigate a bit further to test where the speed gets so lacked down...

EDIT: same problem with an external interrupt (INT0)
EDIT2: found out that disableing cli(); before my setup fexes the INT0 problem - the overflow problem is still there and I have no f**** clue why that is happening...

@techylah

I totally agree to your idea! The sd card interface was always supposed as an animation memory. But I dont understand how to use your code snippet to read it binary.
I think it would do the same job, if I just read a binary file, and instead of an ASCII character I get the byte I want directly (8x Times faster). Is there any way to stream that bytes directly into my array - without the use of counters, comperators, and so on - (that decrease speed extremely)?
But what is a shame,  that I cant find ANY binary level file editor for windows! All "binary" editors show the value only in HEX, DEC, or ASCII! I tried a bunch of them and no one has the function to display it really binary!
Here is how I did it before (reading the ASCII):

Code:

myFile = SD.open("data1.txt")
 if (myFile) {  //does if file exists
    sd_member = 0;  //resets for array mergeing
    sd_position = 0;
    sd_layer = 0;
    while (myFile.available()) {   // reads from the file until there's nothing else in it
      data1[sd_layer][sd_member] = data1[sd_layer][sd_member] << 1;  //shifts it one position to the left
      data1[sd_layer][sd_member] |= (myFile.read() - '0');   //myFile.read acts like a serial communication - all fetched data is ASCII (for this reason -'0' : 48-48 = 0 ; 49-48=1
      sd_position++;  //counts up position
      if (sd_member == 32) {
        sd_member = 0; //resets member that new layer starts being written at the beginning member
        sd_layer++; //counts up layer to acces trough array
      }
      if (sd_position == 16) { //checks if lenth of one member is reached (as for datatype word its 16)
        sd_position = 0; //resets position that new member starts being written at the beginning
        sd_member++; //counts up member to acces trough array
      }
      }
    }
 myFile.close();    // close the file

grz mx

finally got that damn chip to life!
I used many different bootloader and only this one worked:
https://github.com/jmgiacalone/sanguino1284p

its a mix out of many different packages available, and it work very well.
Something strange is, that the 1284p works about 20% faster than my atmega2560 running the same sketch
(i did a shift out demo for an 16bit regsister to steer the leds)
the 1284 does update the 16bit LED display at 13,5khz and the 2560 at 11,6khz (calculated on 50000 cycles each)

now I need to port my current sktches to that baby. Hopefully the flash library will work.
thx for your help!

I have already read the fuse settings with Nick Gammon´s tool -please look my reply #9
Now I tried the same to completely virgin 1284p - same thing happens...

I gonna get mad....

I followed that tutorial to set the 1284 up on a breadboard : http://wiki.wiring.co/wiki/Creating_your_own_Wiring_board_with_a_DIP_atmega1284p_using_Duemilanove
Can it be a problem that i use ceramic disk capacitors?
I dont know what I am doing wrong! - I followed all instruktions properly.
Please tell me what I can do.
Is there any way that I can check if the speeds are identical (I have no osciloscope here)?

BTW.: I would like to buy a Atmel AVR ISP MKii aswell. The genuine from avr costs around 40$ and a rip of clone from the bay around 15$ - can I buy the rip of clone? - or should I better buy the original from AVR direktly?

grz mx

wow thx crossroads - you finally make me feel a little better after that senseless day of trying to get it run.
Now I connected the reset as well, and it at least pops out another error.
It compiles successfull, and avrdude starts to send data (RX flashes) it also gets response from the 1284p, but it never finishes the upload successfull...
it stops at an random point at upload (I think its never the same) and pops me that error:

Code:

avrdude: stk500_paged_write(): (a) protocol error, expect=0x14, resp=0x64
avrdude: Send: V [56] @ [40] . [00] . [00] . [0c]   [20]
avrdude: Recv:
avrdude: stk500_cmd(): programmer is out of sync

maybe there is something wrong with the bootloader? or is there a wrong fuse setting?

grz mx

I am sorry techylan.... I am quite new to all that stuff, and I read numbers (ACII) from an.TXT file on the sd card, and substracted '0' to get the real '0' and '1' out of the .TXT and then I shifted that boolean into a 16 Bit number and stuffed an array fith lots of loops and counters.
I´ve never seen such a method like you mentioned before... Can I anywhere look up for more details how to use that?

But my main problem is still getting the 1284P to life :-(
Today I spent many order with burning different bootloaders and filtering the web for good answeres.
So far, I cant spot any mistakes
-I use a fresh install of the Arduino 1.0 IDE
-I use the Mighty 1284P (https://github.com/maniacbug/mighty-1284p)
-I followed the instruktions on http://maniacbug.wordpress.com/2011/11/27/arduino-on-atmega1284p-4/
-The bootloader led (on PIN 1) also flashes when I press reset.
-I have an 16MHz external crystal with two 22pF tied to ground
-I want to use an Arduino UNO board to upload the sketches (as an serial adapter) (of course I removed the 328P)
   --> I connected the RXD0 to the RX of the arduino UNO board and the TXD0 to the TX

It says "bootloader burn done", but when I want to upload a simple sketch (e.g. "Blink"), I get no response from the 1284P
Here is the log from the AVRDUDE:

Code:

Using Port                    : \\.\COM5
         Using Programmer              : arduino
         Overriding Baud Rate          : 115200
avrdude: Send: 0 [30]   [20]
avrdude: Send: 0 [30]   [20]
avrdude: Send: 0 [30]   [20]
avrdude: Recv:
avrdude: stk500_getsync(): not in sync: resp=0x00

avrdude done.  Thank you.

I have also tried to look if the fuse bits are set properly. - I did this by a tool created by Nick Gammon (http://arduino.cc/forum/index.php?topic=105089.0)

Here is the result:

Code:

Atmega fuse calculator.
Written by Nick Gammon.
Entered programming mode OK.
Signature = 0x1E 0x97 0x05
Processor = ATmega1284P
Flash memory size = 131072
LFuse = 0xFF
HFuse = 0xDE
EFuse = 0xFD
Lock byte = 0xCF
Clock calibration = 0x9D
OCD Enable.............................. [ ]
JTAG Enable............................. [ ]
Enable Serial (ICSP) Programming........ [X]
Watchdog Timer Always On................ [ ]
Preserve EEPROM through chip erase...... [ ]
Boot into bootloader.................... [X]
Divide clock by 8....................... [ ]
Clock output............................ [ ]
Bootloader size: 4096 bytes.
Start-up time: SUT0: [ ]  SUT1: [ ] (see datasheet)
Clock source: low-power crystal.
Brownout detection at: 2.7V.

So far the fuses are set properly.

I hope you guys have an idea....
grz mx

@cr0sh:

Thanks for that advice - safety is always the most important law inhopefully every hacker's basement...
I was wearing protective glasses when spinning my construction up the first time. - after about 3 Minutes of 2000RPM spinning the Arduino (328P) came loose, and flew fast from one end of the room to the other and hit the wall quite strong. - But it is still working (my eye and the 328P)
After my project is finished, I want to place an clear acryllic dome (costs around 50$) over my 3DPOV.

@CrossRoads:

The Atmga 328P was not powerfull enough for the qoals I set anyway - hence a complete redesign is neccassary (your advice comes a little too late ) - thx anyway

The board looks nice, but I already ordered an Sanguino Rip-Off board in the bay (10 bucks or so).
The 1284P arrived yesterday, and I burnt that mighty bootloader.
The burning was successfull, and a led on pin 1 flashed every time I pressed reset - so bootloader should been fine - but i wasnt able to upload a sketch...
it said " avrdude: stk500_getsync():not in sync: resp=0x00 " - I know that this is a common problem - I tried pressing reset before upload and I selcted the proper port and the arduino´s UNO COM Drivers are installed as well (I used an Arduino UNO board without the 328P to upload the sketch).

... So i swapped over to the Sanguino Bootloader - burning was successfull aswell (i think so). But I cant uplad sketches as well. This time it says:
on Arduino 1.0:     error: Arduino.h: No such file or directory
on Arduino r23:     avrdude: AVR Part "atmega1284p" not found. (...and makes a list of all valid AVR Parts)

I dont know what I mistake I would have made...?
Hopefully someone of you genius can help a burning noob out of his problem.

Wow thank you a lot for that hint with the 1284! I didnt knew that chip before, and it seems to be powerfull enough!.
I´ve ordered some and hope I can get em to work. I read many tutorials how to burn the bootloader on that chip, but many people complain about problems.
The wiring bootload should work original on that chip - what would you recomand?

Thx anyway for that great hint!