Print

[lhounsell]

Hi,
I have been building an Astroeq based around an Arduino Mega 2560 but seem to be having some problems.

Setup

Bresser EXOS2 mount (next Revison Of the LXD75)
1.8 Deg Steppers with 2X1 belts on both axis

Motor Driver IC Type - A4988/3
Motor Microstep Level 16uSteps

Right Ascension Axis
Motor Step Angle - 1.8
Motor Gear Ratio - 2
Worm Gear Ration - 144
Goto Rate - 528
Reverse Axis Direction - ON

Declination Axis
Motor Step Angle - 1.8
Motor Gear Ratio - 2
Worm Gear Ration - 144
Goto Rate - 528
Reverse Axis Direction - ON

Calculated RA Settings
Axis Steps/rev 921600
Worm Steps/rev 6400
Sidereal Steps/s 6364
Sidereal Speed 595
Resolution 1.406

Calculated Dec Settings
Axis Steps/rev 921600
Worm Steps/rev 6400
Sidereal Steps/s 6364
Sidereal Speed 595
Resolution 1.406

(1) The mount seems to slew quite accurately, infact more accurate that my Meade 497 with the first couple of stars.
(2) works very well after it has a couple of sync point created with AstroTortilla
(3) For Visual work it would work fine but mostly I would like to concentrate on photography
(3) Tracking seems very strange the image seems to bounce back and forth, maybe 1-3 pixels and this blurs the image. In Stellarium the scope marker move back and forth by 1-2 Arc Seconds.
(4) I cannot get PHP2 to settle down and it seems to be overcompensating no matter what the setting are.

While compiling I noticed the following warnings (is this OK)
_____________________________________________
sketch\AstroEQ.cpp: In function 'int main()':

sketch\AstroEQ.cpp:546:55: warning: deprecated conversion from string constant to 'char*' [-Wwrite-strings]

                     Serial_writeStr(SPI_WELCOME_STRING); //Send version number

                                                       ^

sketch\synta.cpp: In function 'void synta_assembleResponse(char*, char, long unsigned int)':

sketch\synta.cpp:49:45: warning: narrowing conversion of 'responseData' from 'long unsigned int' to 'byte {aka unsigned char}' inside { } [-Wnarrowing]

             Nibbler nibble = { responseData };

                                             ^

sketch\synta.cpp:52:51: warning: narrowing conversion of 'responseData' from 'long unsigned int' to 'unsigned int' inside { } [-Wnarrowing]

             DoubleNibbler nibble = { responseData };

                                                   ^

sketch\synta.cpp: In function 'byte hexToByte(char*)':

sketch\synta.cpp:136:41: warning: narrowing conversion of '(((int)hexToNibbler((* hex))) << 4)' from 'int' to 'byte {aka unsigned char}' inside { } [-Wnarrowing]

     Nibbler high = {hexToNibbler(hex[0])<<4};

                                         ^

Sketch uses 10152 bytes (3%) of program storage space. Maximum is 253952 bytes.
Global variables use 846 bytes (10%) of dynamic memory, leaving 7346 bytes for local variables. Maximum is 8192 bytes.
____________________________________________

I also used some test software to test the delays on the Arduino and got the following results

 delayMicroseconds(20)    : 18.900 us
 delayMicroseconds(80)    : 79.250 us
 delayMicroseconds(90)    : 89.300 us
 delayMicroseconds(100)    : 99.400 us

I also wrote some test software to test the microstepping which seems to show that the A4988's are stepping OK.

I'm not sure if the problem is with the Arduino timing or with my build.

any help on how to track down the problems would be appreciated.

Les

[TCWORLD]

Looking at your numbers, it may be simply that your mount gear ratio is too load. The reported resolution is only 1.41 arcseconds/step - that basically means that for each microstep of the motor, the mount will move over 1.4 arcseconds. Ideally that number needs to be <1, and the lower the better.

The jumping of the cursor in Stellarium is probably due to the fact that it takes a little while for the AstroEQ to report its current position - it is tracking and updating its position correctly internally, but EQMOD only polls for a new position every 100ms or more.

The blurriness/shaking will be because the mount is moving quite a long way with each step - it will move 1.4 arcsec, then stop, then moving another 1.4 arcsec, then stop, and so forth.

If possible I would look in to increasing either the gear ratio or getting some 0.9 degree steppers or both - basically to try and push that resolution number down. Halving the step angle or doubling the gear ratio would halve the number of arcseconds per step.

[lhounsell]

Hi Tom,

Thanks for the very quick reply, nice to see somebody supporting there software.

I have a pair of Nema 17 0.9 motor in the cabinet so I will look at implementing them this week and I will also look at sourcing some over gears.

Grateful for the assistance as I really like EQMOD and you software has made it possible for me to use it.

regards
Les