GPS integration

EOSBandi · Post by **EOSBandi** » Tue May 15, 2012 6:13 am

howardhb wrote:Thanks @EOSBandi!

I will test this in the morning.....

Please confirm, In IMU.pde:
Code: Select all
    // Attitude of the cross product vector GxM
//******************************************EOSBANDI - removed /10 to increase precision of decliniation calc   
    heading = _atan2( EstG.V.X * EstM.V.Z - EstG.V.Z * EstM.V.X , EstG.V.Z * EstM.V.Y - EstG.V.Y * EstM.V.Z  );
//******************************************EOSBANDI  
    //add declination
    //Heading is -180 - 180 had to wrap up accordingly
    heading = heading + MAG_DECLINIATION

   heading = heading /10;               // /10 moved here from above
It seems that only a 10th of the defined magnetic declination is being added to heading, owing to the division by 10, after adding declination?
Is this as you require it?

you are right, nice catch. It sould be
heading = headig + MAG_DECLINIATION * 10;

Katch · Post by **Katch** » Tue May 15, 2012 10:05 am

MelihK wrote:Improved version of EOS Bandit's I2C GPS Code,

Features:
9600 baud standard GPS switching code added, this MTK command switching the GPS module to 115200 baud. You can use same command set with any baudrate to 115200
NMEA sentence configurator MTK command added. It is removing useless NMEA codes from the stream.
LED indicator added. Arduino LED(pin 13) blinking for 3 seconds at start and very short blinks after GPS lock.

@EOS Bandit, Please check it and add into your repository, if they are useful features.

I2C_GPS_NAV.zip

EOS Bandit are these changes implemented in your new code - should it work with Melih's Navigtron without changes?

EOSBandi · Post by **EOSBandi** » Tue May 15, 2012 10:46 am

GPS speed : yes
NMEA sentences config : no, will put there soon, but it's not really an issue
LED indicator: improved thanks to Guru_Florida
* 3 short flash at startup
* blinks once per second if gps is sending data but no position lock yet
* blinks twice fast, then off for a second if gps 2D position is available
* blinks 3 times fast, then off for a second if gps 3D position is available
* or goes on for a second, off for a second, (long pulse) if not getting any NEMA sentences for over 5 seconds

howardhb · Post by **howardhb** » Tue May 15, 2012 10:57 am

Arrrrggghh - Wind and Rain preventing testing... :shock:

In the mean time, if anyone is interested, I've updated LCD.pde so that GPS parameters can be displayed and edited via VT100 Terminal /LCD / OLED.
Just replace LCD.pde with this one. (in the attached zip)

H.

EOSBandi · Post by **EOSBandi** » Tue May 15, 2012 11:02 am

Great, will put it to the repo afternoon

Katch · Post by **Katch** » Tue May 15, 2012 11:13 am

sorry if this is a silly question but why is there an f at the end of the magnetic declination?

can someone check mine

2 degree 57 minutes west = -2.59f ?

howardhb · Post by **howardhb** » Tue May 15, 2012 11:27 am

2 degree 57 minutes west = -2.59f ?

@Katch, the "f" simply means that the #define is a Floating point number.
Yes, -2.59f is correct for declination of: 2 degrees, 57 minutes, West

Katch · Post by **Katch** » Tue May 15, 2012 11:31 am

Ok - next problem

Getting no satellites in the GUI. Navigtron board with the latest i2c code

GPS is green in the GUI
3 i2c errors
at start up the red and blue light on the navigtron flash one sec opposite each other - after a while the red light stops flashing - blue light continues every 1 second.

edit

I've tried this;

Code: Select all

  //Assume 9600 bps
 Serial.begin(9600);
 Serial.print("$PMTK251,115200*1F\r\n");
 delay(1000);
 Serial.end();
 
 Serial.begin(115200);
 Serial.println("$PMTK314,0,1,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0*29");
  delay(1000);//Enable it for 10Hz mode 
  Serial.println("$PMTK300,100,0,0,0,0*2C"); //Enable it for 10Hz mode

but get the same results.

I've also tried putting a test sketch on the arduino before flashing to make sure the GPS is back on 9600.

Edit 2

I'm getting serial data off the FTDI interface at 9600 ...

Katch · Post by **Katch** » Tue May 15, 2012 12:50 pm

winner winner chicken dinner

This version works with Flytron Navigtron and also removes useless nmea sentences.

Code: Select all

  //Assume 9600 bps
  Serial.begin(9600);
  Serial.print("$PMTK251,115200*1F\r\n");
  Serial.end();
  delay(1000);
  Serial.begin(115200);
  Serial.print("$PMTK220,100*2F\r\n");
   delay(1000);
   Serial.println("$PMTK314,0,1,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0*29");
  delay(800);//Enable it for 10Hz mode 
  Serial.println("$PMTK300,100,0,0,0,0*2C"); //Enable it for 10Hz mode

also attached as zip ready to go.

You may need to flash this first;

Code: Select all


/*
  Blink
  Turns on an LED on for one second, then off for one second, repeatedly.
 
  This example code is in the public domain.
 */
 
// Pin 13 has an LED connected on most Arduino boards.
// give it a name:
int led = 13;

// the setup routine runs once when you press reset:
void setup() {                
  // initialize the digital pin as an output.
  pinMode(led, OUTPUT);     
}

// the loop routine runs over and over again forever:
void loop() {
  digitalWrite(led, HIGH);   // turn the LED on (HIGH is the voltage level)
  delay(1000);               // wait for a second
  digitalWrite(led, LOW);    // turn the LED off by making the voltage LOW
  delay(1000);               // wait for a second
}

to make sure the GPS is back on 9600 to receive the mtk commands in setup.

howardhb · Post by **howardhb** » Tue May 15, 2012 1:06 pm

@Katch, your GPS (MTK3329?) cannot understand this:

Code: Select all

Serial.println("$PMTK300,100,0,0,0,0*2C"); //Enable it for 10Hz mode

To set 10Hz, try this:

Code: Select all

Serial.println("$PMTK220,100*2F\r\n");

Katch · Post by **Katch** » Tue May 15, 2012 1:15 pm

That was a line of code suggested by Melih and I couldn't get the GPS working without it.

Yes it's MTK3329

here's a datasheet for the MTK commands http://www.flytron.com/pdf/MTK_commands.pdf

howardhb · Post by **howardhb** » Tue May 15, 2012 1:29 pm

I see... Those instructions are different to the DiyDrones MTK3329.

I'm glad you've got yours working! :mrgreen:

Katch · Post by **Katch** » Tue May 15, 2012 1:33 pm

just wish it wasn't 30mph wind here today.... nevermind

dramida · Post by **dramida** » Tue May 15, 2012 3:25 pm

I found one cause of inconsistent GPS HOLD behaviour. It seems that the GPS readings are not-so consistent. Look to this short clip at gps values as the copter is sitting on the ground.
http://www.youtube.com/watch?v=KTyao-gr_jM
It would be nice to filter out those extremes as we do with giro with a moving average vector.

LenzGr · Post by **LenzGr** » Tue May 15, 2012 3:35 pm

howardhb wrote:I see... Those instructions are different to the DiyDrones MTK3329.

Speaking of which, I too have the MediaTek MT3329 GPS from DIYdrones, which seems to use a custom binary protocol. Is this one supported by the current code? AFAIK it can be configured to use the classic NMEA protocol at 10Hz, but this requires sending commands to the device. Would it make sense to add support for parsing the DIYdrones protocol as well?

EOSBandi · Post by **EOSBandi** » Tue May 15, 2012 3:40 pm

LenzGr wrote:
howardhb wrote:I see... Those instructions are different to the DiyDrones MTK3329.

Speaking of which, I too have the MediaTek MT3329 GPS from DIYdrones, which seems to use a custom binary protocol. Is this one supported by the current code? AFAIK it can be configured to use the classic NMEA protocol at 10Hz, but this requires sending commands to the device. Would it make sense to add support for parsing the DIYdrones protocol as well?

DIYDrones gps is starting with NMEA/38400baud. It is switched to custom binary mode by the arducopter firmware. So you can use it with currect i2cgps code. only start at 38400 baud, and change the refresh rate and the baud rate....

Scotth72 · Post by **Scotth72** » Tue May 15, 2012 4:21 pm

LenzGr wrote:
howardhb wrote:I see... Those instructions are different to the DiyDrones MTK3329.

Speaking of which, I too have the MediaTek MT3329 GPS from DIYdrones, which seems to use a custom binary protocol. Is this one supported by the current code? AFAIK it can be configured to use the classic NMEA protocol at 10Hz, but this requires sending commands to the device. Would it make sense to add support for parsing the DIYdrones protocol as well?

See this thread:
viewtopic.php?f=6&t=1682

LenzGr · Post by **LenzGr** » Wed May 16, 2012 9:36 am

EOSBandi wrote:DIYDrones gps is starting with NMEA/38400baud. It is switched to custom binary mode by the arducopter firmware. So you can use it with currect i2cgps code. only start at 38400 baud, and change the refresh rate and the baud rate....

Ah, good to know! Thanks for this info, will give it a try. I somehow assumed it defaults to the custom binary protocol...
That spares me from having to update the firmware, I guess

howardhb · Post by **howardhb** » Wed May 16, 2012 11:12 am

I must say that my GPS is working far better after flashing the firmware with:

FMP04_Camled_010327_A1.5E_20110118_10Hz_115200.bin

( see previous post from from @Scotth72 )
Just follow the instructions 100%.
You have to load Flashing firmware for the Flashtool program, AND GPS firmware.....

After flashing, I used the latest version of Mini GPS Tool V1.7.1 ( http://www.gtop-tech.com/jsf/download.jsf )
Set COM port to 38400, then press Connect. NB: Then press Ctrl + Alt + S to show the setup menu.
I was then able to set the GPS to output ONLY GGA, GSA and GSV sentences at 115200baud and 10Hz, and SAVE these settings.

So, after doing this, the MTK3329 (DiyDrones) and code from EOSBandi (both serial and I2C) can be used with confidence, at 115kb and 10Hz.

I had a chance, this morning, to test the new GPS Position Hold.
Position Hold definitely works very well. Hovered for 8 minutes in an area of about 1m radius, hands off!
Conditions were calm, almost no wind, but some small gusts, enough to see copter correcting!
Need to play some more with RTH PID's - overshoots a bit +- 10m, getting progressively worse.
I think lowering NavR P and PosR P will be the place to start.

H.

LenzGr · Post by **LenzGr** » Wed May 16, 2012 11:22 am

Awesome. I'll give this a try, thanks a lot!

jevermeister · Post by **jevermeister** » Wed May 16, 2012 11:55 am

howardhb wrote:I must say that my GPS is working far better after flashing the firmware with:

FMP04_Camled_010327_A1.5E_20110118_10Hz_115200.bin
( see previous post from from @Scotth72 )
Just follow the instructions 100%.
You have to load Flashing firmware for the Flashtool program, AND GPS firmware.....

After flashing, I used the latest version of Mini GPS Tool V1.7.1 ( http://www.gtop-tech.com/jsf/download.jsf )
Set COM port to 38400, then press Connect. NB: Then press Ctrl + Alt + S to show the setup menu.
I was then able to set the GPS to output ONLY GGA, GSA and GSV sentences at 115200baud and 10Hz, and SAVE these settings.

So, after doing this, the MTK3329 (DiyDrones) and code from EOSBandi (both serial and I2C) can be used with confidence, at 115kb and 10Hz.

I had a chance, this morning, to test the new GPS Position Hold.
Position Hold definitely works very well. Hovered for 8 minutes in an area of about 1m radius, hands off!
Conditions were calm, almost no wind, but some small gusts, enough to see copter correcting!
Need to play some more with RTH PID's - overshoots a bit +- 10m, getting progressively worse.
I think lowering NavR P and PosR P will be the place to start.

H.

hey,

Is this neccesary to run the GPS with the nu code what is thhe difference?what is neded to flash the fmp?

thank ya
nils

EOSBandi · Post by **EOSBandi** » Wed May 16, 2012 12:21 pm

Guys, I found a serious issue with the PID settings, pushed an update please use it.
(I turned out that teh PIDITEMS define is not used in the definitions of the PID arrays

)

The updated gps firmware adds defuault 10Hz and 115200bps settings, plus allows you to save your desired settings into the flash on the gps module. Which removes the burden of setting the gps module on the fly (which is a kind a tricky and unpredictable especially with navigatron

)

LuFa · Post by **LuFa** » Wed May 16, 2012 12:26 pm

only at the i2c gps code ?? or also at Serial GPS Code ?

EOSBandi · Post by **EOSBandi** » Wed May 16, 2012 12:29 pm

howardhb wrote:
I had a chance, this morning, to test the new GPS Position Hold.
Position Hold definitely works very well. Hovered for 8 minutes in an area of about 1m radius, hands off!
Conditions were calm, almost no wind, but some small gusts, enough to see copter correcting!
Need to play some more with RTH PID's - overshoots a bit +- 10m, getting progressively worse.
I think lowering NavR P and PosR P will be the place to start.

H.

Hi. try to increase the D terms... i flew with these last time... and got a quite OK RTH...

#define POSHOLD_P .11
#define POSHOLD_I 0.0
#define POSHOLD_RATE_P 1.4
#define POSHOLD_RATE_I 0.2
#define POSHOLD_RATE_D 0.018
#define NAV_P 1.4
#define NAV_I 0.22
#define NAV_D 0.008

You also can lower the navigational speed max (at the moment it's not wired out to gui, so you have to change it in the i2c_gps_nav config.h NAV_SPEED_MAX to 300...)

EOSBandi · Post by **EOSBandi** » Wed May 16, 2012 12:29 pm

LuFa wrote:only at the i2c gps code ?? or also at Serial GPS Code ?

Multiwii base code..

jevermeister · Post by **jevermeister** » Wed May 16, 2012 12:35 pm

Argh I so want to test your code mate, but I broke my hand and can not use my tx :-(

Are u in contact with alex, will it be in the trunk?

nils

LuFa · Post by **LuFa** » Wed May 16, 2012 12:41 pm

EOSBandi wrote:
LuFa wrote:only at the i2c gps code ?? or also at Serial GPS Code ?

Multiwii base code..

ok

because i have only see a update from the i2c code

howardhb · Post by **howardhb** » Wed May 16, 2012 1:32 pm

#define POSHOLD_P .11
#define POSHOLD_I 0.0
#define POSHOLD_RATE_P 1.4
#define POSHOLD_RATE_I 0.2
#define POSHOLD_RATE_D 0.018
#define NAV_P 1.4
#define NAV_I 0.22
#define NAV_D 0.008

Thanks! Will fly again this afternoon. (un-characteristically perfect, windless conditions here! :mrgreen:

)

If you wouldn't mind, please push my updated LCD / OLED ... (attached zip)

EOSBandi · Post by **EOSBandi** » Wed May 16, 2012 1:33 pm

LuFa wrote:
EOSBandi wrote:
LuFa wrote:only at the i2c gps code ?? or also at Serial GPS Code ?

Multiwii base code..

ok

because i have only see a update from the i2c code

http://code.google.com/p/i2c-gps-nav/source/detail?r=22
Log message
A serious issue with PID settings identified and corrected.
(PID array definition did not used PIDITEMS define... )
Without this fix r20 and r21 is unflyable!!!!!

Modify /trunk/I2C_GPS_NAV/I2C_GPS_NAV-MultiwiiDev-NAV/MultiWii/MultiWii.ino
Modify /trunk/I2C_GPS_NAV/I2C_GPS_NAV-MultiwiiDev-NAV/MultiWii/Serial.ino
Modify /trunk/I2C_GPS_NAV/I2C_GPS_NAV-MultiwiiDev-NAV/MultiWiiConf/MultiWiiConf.pde

LuFa · Post by **LuFa** » Wed May 16, 2012 1:50 pm

ah ok , you have also the Serial GPS code move to the i2c gps project page

because the latest version wich i have use is r10 . Have r10 the same proplem ?

if i understand it right , r22 is for i2c and Serial GPS ?

Katch · Post by **Katch** » Wed May 16, 2012 2:05 pm

r22 error while compiling... line 435 multiwii.ino

debug3=get_free_memory();

compiler message

MultiWii.cpp: In function 'void annexCode()':
MultiWii:434: error: 'get_free_memory' was not declared in this scope

Edit - ok something seriously wrong with r22

using FreeIMU 0.4.3

removed the debug line above and it will compile and upload

Once done the copter no longer has any pitch or roll control - the channels change in the GUI but do not change the motor speed.

Something very broken - do not try to fly this r22 without verifying you definitely have pitch and roll control - I don't

back to 21 and everything works again.

EDIT 2 - seems to be strangely linked to level mode being active. need to do more tests.

IMPORTANT EDIT

Seems to be linked to the way my i2c gps is attached to my Paris v4 - I have moved to my MWC with FreeIMU 0.3.5ms and its behaving again.

dr.tom · Post by **dr.tom** » Wed May 16, 2012 3:17 pm

Katch wrote:Once done the copter no longer has any pitch or roll control - the channels change in the GUI but do not change the motor speed.

Something very broken - do not try to fly this r22 without verifying you definitely have pitch and roll control - I don't

how is your rate setting set?
is it left @ 0 ?

in gui, it seems to work OK,
stick input(pitch/roll) is shown in motor action when armed. in gyro and level mode, both.

Katch · Post by **Katch** » Wed May 16, 2012 3:44 pm

Ok - just flown 2 pack on my MWC 0.3.5ms with I2C GPS

Still a little windy so not getting great results - but getting results nonetheless.

10mph winds

Return to home very definitely brought it back a couple of times and stopped nicely overhead.

Pos Hold - seems to either wander off or circle larger and larger circles until I have to take control.

Still not sure why my Paris v4 board was playing up but I do have my IMU on the LV side of the LLC and the GPS on the HV side. Something is messing with the i2c data along the line somewhere as it is only when the GPS board is attached that I have this lack of control.

EOSBandi · Post by **EOSBandi** » Wed May 16, 2012 3:54 pm

Katch wrote:Ok - just flown 2 pack on my MWC 0.3.5ms with I2C GPS

Still a little windy so not getting great results - but getting results nonetheless.

10mph winds

Return to home very definitely brought it back a couple of times and stopped nicely overhead.

Pos Hold - seems to either wander off or circle larger and larger circles until I have to take control.

So it's time to start with the settings. Try increase POSHOLD_RATE_D, or decrease POSHOLD_P and POSHOLD_RATE_P. If in RTH it comes back and stops then it means your settings are not far from OK.

Katch · Post by **Katch** » Wed May 16, 2012 4:46 pm

waiting for my packs to charge - will go out after dinner and hopefully less wind.

EOSBandi · Post by **EOSBandi** » Wed May 16, 2012 4:55 pm

I pushed meanwhile r24. No changes exept I removed the erronous line and added Howard's lcd.ino modifications. (Thanks Howard !) and compiled the MultiWiiConf for Mac too.

Gaijin · Post by **Gaijin** » Wed May 16, 2012 6:25 pm

EosBandi,

2 items of feedback:

1) With this release my pitch and roll inputs to the FC no longer have any effect even though the board will arm and hover just fine , the TX inputs appear correct in the GUI, input provided by CPPM on a futaba mapping, worked okay on your release from earlier this week.

2) The GUI will not connect to Com 15 via bluetooth serial, it just doesn't list it, again the last iteration was fine is this regard, the side effect is I can't cal the mag with the motors spinning and I can't check the GPS info whilst it is powered by battery.

I2C errors whilst powered by USB however are down to 0 from 6 on Quadrino Zoom /Flytron Navigatron setup

Katch · Post by **Katch** » Wed May 16, 2012 6:27 pm

That is the same set of observations I found with my Paris v4 and GY_86 IMU

The problems did not persist when I switched to my MWC board - FreeIMU 0.3.5ms all running of 5v and HV i2c lines

Katch · Post by **Katch** » Wed May 16, 2012 7:15 pm

Increased PosR from 1.4 to 1.7

hold seems to be a smaller circle now.

There still seems to be some level of randomness in whether RTH or POS Hold work well or not. I can fly to a similar location 4 or 5 times and engage POS Hold and get different results ranging from a good hold to heading off in some random direction.

I've also notice a pronounced kick of the motors when POS Hold is engaged which tends to throw the quad off its hold point and starts the pendulum effect circles.

RTH always brings it back and always slows as it approaches - after that its the same as POS Hold - sometimes good sometimes bad.

I also noticed that altitude hold is not great in either case - the quad often gains or looses altitude beyond what is acceptable. I've had it climb 30 or more meters and in another instance loose altitude from 20m to the ground.

dr.tom · Post by **dr.tom** » Wed May 16, 2012 7:17 pm

happy for you that can try it 8-)

I get tons of I2C errors, gps is green in GUI, but no gps data is shown, all zeros.

Crius SE + Navigatron
all flashed ok, with Katch's blink code first, it blinks ON/OFF
then with latest EosBandi's code+Katch's patch for navigatron@10Hz,
Navigatron blinks 3times per sec after 10sec from powerup = 3D fix, so that works ok i guess.
(without patch it blinks 1sec ON/OFF long= 'no nmea', because of low baud rate i guess)

is there some conflict with BMA180? 0x40 address is same for gps and bma,
i'm no expert, just found that by search function when looking what addresses are used for what... maybe I'm wrong

stock navigatron firmware works ok on previous MWC releases (dev20120504), flight tested with gps RTH OK,
so hardware part + connections are ok.

thanks for help

EOSBandi · Post by **EOSBandi** » Wed May 16, 2012 7:47 pm

It's definitely not an address issue.
i2cgps address is 0x40 BMA180 is 0x80, and i flown today with a Crius board (not an SE).
Did you guys disabled LCD_CONF and VBATT Code in config.h ?
I'm afraid that even with the short i2cgps code we are running out of ram memory.
The ONLY change between r21 and r24 is that the Pid array has a correct size now, so I'm baffled.
Could you send to me your config file ?
EOSBandi

copterrichie · Post by **copterrichie** » Wed May 16, 2012 9:01 pm

wilco1967 wrote:
What I did observe was it might not always come home exactly to the right spot, but perhaps 10..20 meters from it downwind.I guess the wind is to blame for that.

Anyway, not perfect, but still good enough to bring it back to me if I loose orrientation.... and it is way cool to leave it hanging there without touching the controls.

Thanks you very much for the hard work, EOSBandi

Now lets see if I can get 'approval' from the wife for a 10 Hz GPS

Wilco

I use an old 4800 baud Microsoft GPS unit with my ODS and it does Drift but it is reliable. I hope to purchase a new GPS with all of the great features but for testing and development, this one that I have does just great. What I found to be true with this unit is, allow it to lock onto as many satellites as possible before fixing the home position.

Gaijin · Post by **Gaijin** » Wed May 16, 2012 9:22 pm

EOSBandi wrote:It's definitely not an address issue.
i2cgps address is 0x40 BMA180 is 0x80, and i flown today with a Crius board (not an SE).
Did you guys disabled LCD_CONF and VBATT Code in config.h ?
I'm afraid that even with the short i2cgps code we are running out of ram memory.
The ONLY change between r21 and r24 is that the Pid array has a correct size now, so I'm baffled.
Could you send to me your config file ?
EOSBandi

Sure, here it is for the Mwii portion, the I2C config is stock, no changes made

http://www.filedropper.com/config_7

Is the PID array shared for the input routines perhaps?

ApoC · Post by **ApoC** » Wed May 16, 2012 9:55 pm

Where i have the "Arduino-twi-lib-patch" to put in? Found nothing where it have to copy.

Katch · Post by **Katch** » Wed May 16, 2012 10:18 pm

ApoC wrote:Where i have the "Arduino-twi-lib-patch" to put in? Found nothing where it have to copy.

in the wire lib

arduino-1.0\libraries\Wire\utility

ApoC · Post by **ApoC** » Wed May 16, 2012 10:47 pm

Damn, now where i see my own text - i see it...sorry guys....

Another Question: Wich Modes i have to mix?

I read, that i have to switch in the MAG with the GPS Hold, that it works. Maybe the Baro? Or would the height from the GPS used?

Wich Modes have switchd togheter?

djrm · Post by **djrm** » Wed May 16, 2012 10:54 pm

I have been following this most interesting development, thank you eosbandi.

I have installed the latest (r24) software onto my 8MHz pro mini 328 satnav board. it appears to be working well although I cannot attempt to fly it just yet.

I found a small problem with the led blinking, it would always blink slowly as if there was no 3d lock. I found there is a problem on my system in the way the status is checked for the blinking function. I have modified it like this and it now appears to work correctley:

Code: Select all

    if(_statusled_blinks==0) {
      if(i2c_dataset.status.gps3dfix /*== 1*/)
        _statusled_blinks=3;
      else if(i2c_dataset.status.gps2dfix /*== 1*/)
        _statusled_blinks=2;
      else
        _statusled_blinks=1;      
    }

There are other places in the software where the gpsNfix is checked == 1, I expect they too do not always evaluate as expected. I think the test would be better written as gps3dfix != 0 rather than gps3dfix ==1.

In the archive for the r24 software there is both the old and new versions of the i2c gps nav sketch, the one in the root of the archive looks like it is the correct one, a little confusing.

Keep up the good work, David.

UPDATE ...
The problem still sometimes exists with my modification, I now think the cause must be the $GPGSA frames are not being properley decoded in my arduino. If I reset the board it starts to flash 3 short but then reverts to 1 after a while. Oh well, I'll keep looking.

SOLUTION
It appears that if there are less than 5 satellites used for the lock then the slow flash mode is invoked, I only have 4 in view from my desk.

EOSBandi · Post by **EOSBandi** » Thu May 17, 2012 7:19 am

Morning,
It was a long night, finally I think I found the issue. When changed the PID array size I did not increased the EEPROM version number. So if somebody uploaded a sketch over r22 it messed up the config values, especially calibration and trim values for Acc. It worked on boards which did not had r22 or earlier nav version.
(current multiwii dev trunk has EEPROM version 154, nav code had 155 (now 156)
I pushed r26 to the repo, but a simple reset from the gui will fix it too.

EOSBandi · Post by **EOSBandi** » Thu May 17, 2012 7:26 am

Hi !

In the archive for the r24 software there is both the old and new versions of the i2c gps nav sketch, the one in the root of the archive looks like it is the correct one, a little confusing.

Agree, i fixed this in r26, i2c_gps_nav sketch now lives in it's own folder. I also separated the defines for easier reading.

SOLUTION
It appears that if there are less than 5 satellites used for the lock then the slow flash mode is invoked, I only have 4 in view from my desk.

It was intentional since 5sats are the bare minimum to fly with.

ApoC · Post by **ApoC** » Thu May 17, 2012 10:12 am

After flashing your last Versions, i can see, that the LED is not blinking, when i have satfix. So what i have to do, to get it working?