Onstep First testing question #EQ5


Hemendra Rawat
 

Hi All,

I got my Onstep (STM32 black pill, version 4.24) up and running on my Skyview Pro (EQ5) today. I'm using the skysafari and Onstep android app to test it. Here's what I did.

I opened the android app and setup home position, time and location. Then I used sky safari app to test the goto feature. I tested direction buttons to move the scope to East, West, North and South and it seems to work correctly. My only problem is when I do a goto to an object which is close to horizon, the RA axis moves in the correct direction but the DEC axis is reversed (telescope points up instead of pointing down towards horizon). The DEC axis seems to work fine when I use North(telescope points up) and South(telescope points down) Guiding buttons so I don't think the axis motor is reversed. Does anyone know what can be the reason for such a behavior? I'm using sky safari Pro on android.

Question 2: Can anyone please explain the difference between Home and Park option in the android app. I'm a newbie.  

Regards,
Hemendra


Drew 🔭📷🚴‍♂️
 

Most likely your DEC axis is reversed. Understanding its motion when manually guiding takes awhile to understand. The basic rule is that it always goes North when you press the N button. However, when you start guiding (and when it crosses the meridian during any other guiding) it will head South as you continue to hold the button. Let go of the button and press N again and it will appear to reverse motion. Actually it is just heading North as intended.

To test your motion guide the telescope DEC axis away from the starting NCP position. Now press N and see if it heads in the right direction. I suspect it will not.

The home position is the start or zero position OnStep uses for its calculations and alignment. As such it contains more information than just those coordinates. Resetting it also removes the alignment data. This is important for mobile setups.

The Park position is just a location. It is intended for fixed setups such as a pier mount in an observatory. Celestron called it the "hibernate" position.


Hemendra Rawat
 

Hi All,

I reversed the DEC motor and now my telescope at least point towards Sky. I tested my mount out in the field yesterday and pointing accuracy isn't good at all. So I think I need some more help. Here's what I did:

1) Did Polar alignment.
2) Initialized Onstep (Uploaded Date/Time and Location)
3) Tried a 2 Star Align:
                      Star 1: Selected Dubhe in Ursa Major -> My telescope moved in the right direction but way off the target. I used guide buttons in the alignment window to make it point exactly towards Dubhe. Long pressed the align button.
                      Star 2: Selected Vega -> Again telescope moved in the right direction but again it way off the target. It pointed somewhere right of Polaris (RA and DEC both axis were off the target). Again used guide button to correct the alignment and long press the align button.

4) I then tried finding M13 to test the alignment and the telescope wasn't even in the ballpark location :(

How do I fix this issue? I have an Orion Skyview Pro Eq mount, STM32 black pill, NEMA steppers with 400 Steps/0.9 degree, 24V DC supply. For both axis, motors are connected to mount in 1:2.5 ratio(16T GT2 pulley on motors and 40GT2 pulley on mount drive. Belts are not slipping. Attached are the config details.


philippe.larrieux@...
 

I’ll let the experts answer, but in the meanwhile, you have AXIS1_STEPS_PER_WORMROT set to 32000, enabling PEC processing. It should be set  to 0 except if you have pec sensor. Other values look good.

 

Regarding align, at least 3 stars are recommended to get a good model as far as I know.  And Dubhe is probably not the best choice, too close to polar.

 

 

From: main@onstep.groups.io <main@onstep.groups.io> On Behalf Of Hemendra Rawat
Sent: mardi 8 juin 2021 09:07
To: main@onstep.groups.io
Subject: Re: [onstep] Onstep First testing question #EQ5

 

Hi All,

I reversed the DEC motor and now my telescope at least point towards Sky. I tested my mount out in the field yesterday and pointing accuracy isn't good at all. So I think I need some more help. Here's what I did:

1) Did Polar alignment.
2) Initialized Onstep (Uploaded Date/Time and Location)
3) Tried a 2 Star Align:
                      Star 1: Selected Dubhe in Ursa Major -> My telescope moved in the right direction but way off the target. I used guide buttons in the alignment window to make it point exactly towards Dubhe. Long pressed the align button.
                      Star 2: Selected Vega -> Again telescope moved in the right direction but again it way off the target. It pointed somewhere right of Polaris (RA and DEC both axis were off the target). Again used guide button to correct the alignment and long press the align button.

4) I then tried finding M13 to test the alignment and the telescope wasn't even in the ballpark location :(

How do I fix this issue? I have an Orion Skyview Pro Eq mount, STM32 black pill, NEMA steppers with 400 Steps/0.9 degree, 24V DC supply. For both axis, motors are connected to mount in 1:2.5 ratio(16T GT2 pulley on motors and 40GT2 pulley on mount drive. Belts are not slipping. Attached are the config details.


Dave Schwartz
 

If this was the very first time you set the date/time/location, was that the order you did them in (step 2)? If so, you need to do the location first - especially the UTC Offset (using the app on the phone is the most reliable way to get this right) - because OnStep uses the UTC offset to transform the local mean time that is uploaded into the UTC time which is actually used internally and used to initialize the RTC. If the UTC offset is wrong, and thus the UTC, there will be large pointing errors.

Are the J1 and J2 jumpers for the driver type installed properly? The LV8729 drivers actually do not require these jumpers to be installed (but it doesn't hurt if J1 is on the right two pins and J2 is on the left two pins). If they are on the wrong pins, that will interfere with OnStep's setting of the microstep mode and the motors will travel either more or less than desired.

Please attach your Config.h file to your reply. Its all well and good that we see the spreadsheet and the configurator page but seeing the actual parameter values in the Config.h file is the ultimate truth.

With the STM32 Blue Pill PCB and LV8729 drivers, you can set MICROSTEPS_GOTO to a value smaller than MICROSTEPS and you will get faster goto performance. With MICROSTEPS_GOTO OFF, it will be trying to do everything in 32 microsteps and this limits the motor speed due to the rate at which stepping can be done.

On 2021-06-08 3:07 a.m., Hemendra Rawat wrote:
Hi All,

I reversed the DEC motor and now my telescope at least point towards Sky. I tested my mount out in the field yesterday and pointing accuracy isn't good at all. So I think I need some more help. Here's what I did:

1) Did Polar alignment.
2) Initialized Onstep (Uploaded Date/Time and Location)
3) Tried a 2 Star Align:
                      Star 1: Selected Dubhe in Ursa Major -> My telescope moved in the right direction but way off the target. I used guide buttons in the alignment window to make it point exactly towards Dubhe. Long pressed the align button.
                      Star 2: Selected Vega -> Again telescope moved in the right direction but again it way off the target. It pointed somewhere right of Polaris (RA and DEC both axis were off the target). Again used guide button to correct the alignment and long press the align button.

4) I then tried finding M13 to test the alignment and the telescope wasn't even in the ballpark location :(

How do I fix this issue? I have an Orion Skyview Pro Eq mount, STM32 black pill, NEMA steppers with 400 Steps/0.9 degree, 24V DC supply. For both axis, motors are connected to mount in 1:2.5 ratio(16T GT2 pulley on motors and 40GT2 pulley on mount drive. Belts are not slipping. Attached are the config details.


Khalid Baheyeldin
 

Hemendra,

Looking at tnut's conversion in the showcase page, 144T is right for the SkyView Pro.

The first star is usually off. I usually have a degree or two due to polar alignment.
However, the 2nd star should be in the field of view, if everything is right.

As Phillipe said: do not use stars close to the celestial pole. You need at least 3 stars, and
they have to be in both sides of the meridian.

But if your 2nd star is not in the field of view, then you have another issue somewhere.

Here are some possibilities of the source of errors:

- The motors are actually 200 step and not 400 step, but the configuration does not match.
In this case, the mount will overshoot the target (go beyond it).

- The motors are stalling during the slew (they have a high pitched noise with no movement).
If that is the case, then you need to adjust Vref properly, and maybe change the microstepping
What drivers are you using and how are they setup?


Drew 🔭📷🚴‍♂️
 

On Tue, Jun 8, 2021 at 09:57 AM, Khalid Baheyeldin wrote:
The motors are stalling during the slew
Also, let's not forget mechanical adjustment. The mount must be mechanically smooth when rotated by hand. A sticky place or section will cause alignment issues. Even more important is that the stepper motors we use are generally much more powerful than those on the original drive. They are capable of doing mechanical damage that the old ones could not. Now is the time to resolve these problems.


Hemendra Rawat
 

Attached is the updated config.h file. It has the following changes

1) Disabled PEC by setting "Steps Per Worm Wheel Rotation" to 0
2) Enabled MICROSTEPS_GOTO . It is set to 16 now.   (Didn't see any improvements in slewing speed)

I still see that the mount doesn't properly point to the target. This time I used Vega for the first star and then pointed it towards arcturus. It is off by at least 15-20 degrees in both the axis. 

Some extra information that people asked previously.

1) Confirmed that motors are 400 steps 0.9 degree Nema 17. 
2) Mount is mechanically stable. I can move it smoothly with hands or slow motion knobs.
3) Jumpers J1 and J2 are installed.
4) The motors have a humming sound. There also make noise but they always move when I press the guide buttons. I have not seen them jamming.
5) Checked the belts they are not loose.
6) I'm using Big Tree Tech's LV8729 drivers from AliExpress. The micro-step is set to 32 in the config.h (Do I need to do something else to setup micro-steps)
7) In the app I first set observing site and then set time and then go to align.

Please advise how to debug this issue?

Regards,
Hemendra


Khalid Baheyeldin
 

On Wed, Jun 9, 2021 at 03:19 AM, Hemendra Rawat wrote:
4) The motors have a humming sound.
When tracking, a humming sound is normal. Even if it pulsates a bit, still normal.
There also make noise but they always move when I press the guide buttons. I have not seen them jamming.
Guiding is different from slewing. When you are guiding, the rate of pulsing the motor is still low.
When slewing, OnStep pulses the motor much faster in a gradual fashion over a few degrees, until it reaches the slewing speed.
That is when stalling usually happens.
The motors will stop moving, and there will be a high pitched sound.

6) I'm using Big Tree Tech's LV8729 drivers from AliExpress. The micro-step is set to 32 in the config.h (Do I need to do something else to setup micro-steps)
What did you set Vref to?
What is the maximum current for your motors?


John Petterson
 

Hemendra,

Let’s start with a couple of very basic things that everyone assumes you got right but could be causing these issues.  I know these are very basic things, and I also assume none are the underlying problem but it is good to verify them.

First, very carefully balance the telescope and mount, both in RA and DEC axis. Stepper motors (and the drive train in general) will probably not be damaged by trying to operate out of balance, but the stepper motor will miss steps or the belt will slip and not move the mount as far as expected when the resistance to motion is too great.  You said you can move the mount with slow motion controls.  Is this easy to do, or are the knobs providing resistance when you try to turn them?  It should turn with almost no resistance - like turning the volume knob up on a TV or radio,  Any resistance to movement could mean that the mount needs to be cleaned and greased, or that your balance is off, or that your worm drive mechanism needs adjustment, or (if the resistance gets higher and over as you rotate the knob) the drive gears may be damaged.

Just to confirm, the smaller pulley (the 16 tooth one) is on the motor and the bigger one (40 tooth) is on the shaft connected to the worm gear on the mount.  And the belt should be pulled tight on the pulleys so you cannot press it in or out in the center between the pulleys more than a mm or two.  Any looser than that could allow it to slip over the teeth.

(Assuming you are in the northern hemisphere,)  if you hold the SOUTH button for many seconds, does the mount swing the telescope to point below the horizon in the southeast or southwest?  If instead it rotates around the pole then you have the wires for the DEC and RA motors connected to the wrong motors.  I have done this, it is confusing when it happens because often the mount will move in sort of the right places but be off by a bit.  It becomes obvious when you move one axis at a time, but when you do a slew to start your alignment it is hard to see.

John


Hemendra Rawat
 

Khalid,

My stepper motors are 400 steps, 0.9A. I have set Vref to 0.5V

John,

I will test the mechanics of my mount tonight and let you know.


Regards,
Hemendra


On Wed, Jun 9, 2021, 10:24 AM John Petterson <j.petterson@...> wrote:

Hemendra,

Let’s start with a couple of very basic things that everyone assumes you got right but could be causing these issues.  I know these are very basic things, and I also assume none are the underlying problem but it is good to verify them.

First, very carefully balance the telescope and mount, both in RA and DEC axis. Stepper motors (and the drive train in general) will probably not be damaged by trying to operate out of balance, but the stepper motor will miss steps or the belt will slip and not move the mount as far as expected when the resistance to motion is too great.  You said you can move the mount with slow motion controls.  Is this easy to do, or are the knobs providing resistance when you try to turn them?  It should turn with almost no resistance - like turning the volume knob up on a TV or radio,  Any resistance to movement could mean that the mount needs to be cleaned and greased, or that your balance is off, or that your worm drive mechanism needs adjustment, or (if the resistance gets higher and over as you rotate the knob) the drive gears may be damaged.

Just to confirm, the smaller pulley (the 16 tooth one) is on the motor and the bigger one (40 tooth) is on the shaft connected to the worm gear on the mount.  And the belt should be pulled tight on the pulleys so you cannot press it in or out in the center between the pulleys more than a mm or two.  Any looser than that could allow it to slip over the teeth.

(Assuming you are in the northern hemisphere,)  if you hold the SOUTH button for many seconds, does the mount swing the telescope to point below the horizon in the southeast or southwest?  If instead it rotates around the pole then you have the wires for the DEC and RA motors connected to the wrong motors.  I have done this, it is confusing when it happens because often the mount will move in sort of the right places but be off by a bit.  It becomes obvious when you move one axis at a time, but when you do a slew to start your alignment it is hard to see.

John


Hemendra Rawat
 

Hello John,

Please find the answers to your questions below:

First, very carefully balance the telescope and mount, both in RA and DEC axis. Stepper motors (and the drive train in general) will probably not be damaged by trying to operate out of balance, but the stepper motor will miss steps or the belt will slip and not move the mount as far as expected when the resistance to motion is too great.  You said you can move the mount with slow motion controls.  Is this easy to do, or are the knobs providing resistance when you try to turn them?  It should turn with almost no resistance - like turning the volume knob up on a TV or radio,  Any resistance to movement could mean that the mount needs to be cleaned and greased, or that your balance is off, or that your worm drive mechanism needs adjustment, or (if the resistance gets higher and over as you rotate the knob) the drive gears may be damaged.

>> The mount is few years old but in good condition. The DEC slow motion knob is very smooth (no resistance). The RA slow motion knob is relatively less smooth (compared to DEC axis) but resistance is small. I doubt this is the issue here.

Just to confirm, the smaller pulley (the 16 tooth one) is on the motor and the bigger one (40 tooth) is on the shaft connected to the worm gear on the mount.  And the belt should be pulled tight on the pulleys so you cannot press it in or out in the center between the pulleys more than a mm or two.  Any looser than that could allow it to slip over the teeth.

>> Yes 16T GT2 pulley is on motor shaft and 40T GT2 pulley is on the worm gear shaft. The RA belt is tight as you described. The DEC belt was a bit loose so I replaced it with a smaller belt. It's now tight but I didn't see any improvements in pointing accuracy. 

(Assuming you are in the northern hemisphere,)  if you hold the SOUTH button for many seconds, does the mount swing the telescope to point below the horizon in the southeast or southwest?  If instead it rotates around the pole then you have the wires for the DEC and RA motors connected to the wrong motors.  I have done this, it is confusing when it happens because often the mount will move in sort of the right places but be off by a bit.  It becomes obvious when you move one axis at a time, but when you do a slew to start your alignment it is hard to see.

>> I'm in the northern hemisphere. At home position (pointing towards Polaris) when I press the SOUTH button, DEC axis moves towards right (West in terms of celestial coordinates). Pressing NORTH button moves the DEC axis back towards Polaris. Pressing EAST moves the RA axis towards left (celestial East) and pressing West moves RA axis towards right(celestial West). 

My motors are 400 steps 0.9 degree. I have set the Vref to 0.5V for both the axis. I'm powering Onstep with a 24V 3A power supply. I'm using LV8729 stepper drivers. In 10-15 minutes of testing I didn't see steppers or the stepper drivers getting very hot.

What's the best way to validate that Onstep has correct time, time zone and location information? Based on my observation, Onstep always moves in the right direction but for some reason it's undershooting the target. How can I debug this?

Regards,

Hemendra





Khalid Baheyeldin
 

On Thu, Jun 10, 2021 at 02:56 AM, Hemendra Rawat wrote:
My motors are 400 steps 0.9 degree. I have set the Vref to 0.5V for both the axis. I'm powering Onstep with a 24V 3A power supply. I'm using LV8729 stepper drivers.
On the underside of the driver, there are two resistors. They will be marked R220 or R100.
Let us know which one yours have (check both drivers).

What's the best way to validate that Onstep has correct time, time zone and location information?
You have WiFi. So go to http://192.168.0.1/ and on the main page you will see the coordinates, UTC time, as well as the Local Star Time (LST). Compare the LST to what is on this site after you enter the correct coordinates.

Based on my observation, Onstep always moves in the right direction but for some reason it's undershooting the target. How can I debug this?
Undershooting, and elimination of other reasons, points to stalling.
Can you make a video showing the scope from power up until the end of the slew for the first stall?
Upload it Youtube (do not attach), so we can hear what is going on.
 


Hemendra Rawat
 

Hi Khalid,

Both my Stepper drivers have R220 written on both the resistors.

Here is the link to the video of my mount in action.


I can make another one if details aren't very clear.

Regards,
Hemendra

On Thu, Jun 10, 2021, 9:56 AM Khalid Baheyeldin <kbahey@...> wrote:
On Thu, Jun 10, 2021 at 02:56 AM, Hemendra Rawat wrote:
My motors are 400 steps 0.9 degree. I have set the Vref to 0.5V for both the axis. I'm powering Onstep with a 24V 3A power supply. I'm using LV8729 stepper drivers.
On the underside of the driver, there are two resistors. They will be marked R220 or R100.
Let us know which one yours have (check both drivers).

What's the best way to validate that Onstep has correct time, time zone and location information?
You have WiFi. So go to http://192.168.0.1/ and on the main page you will see the coordinates, UTC time, as well as the Local Star Time (LST). Compare the LST to what is on this site after you enter the correct coordinates.

Based on my observation, Onstep always moves in the right direction but for some reason it's undershooting the target. How can I debug this?
Undershooting, and elimination of other reasons, points to stalling.
Can you make a video showing the scope from power up until the end of the slew for the first stall?
Upload it Youtube (do not attach), so we can hear what is going on.
 


Khalid Baheyeldin
 

On Fri, Jun 11, 2021 at 01:24 AM, Hemendra Rawat wrote:
Both my Stepper drivers have R220 written on both the resistors.
Let us do the math.

Your motors are maximum 0.9A current.
We want to run them at 50% (to start with).

So 0.9 X 1.41 X 0.5 = 0.635A

The LV8729 Vref formula is:

0.635 / 1.1 = 0.577 V

So you need to set your vref to 0.577 V

Then, change the following in your Config.h file:

#define STEP_WAVE_FORM             PULSE // allows faster pulsing during slewing (60% more)

Then disable the different microstepping in case that is stalling (see below).

#define AXIS1_DRIVER_MICROSTEPS_GOTO  OFF
#define AXIS2_DRIVER_MICROSTEPS_GOTO  OFF

Here is the link to the video of my mount in action.
 
 
I can make another one if details aren't very clear.
In your video the mount overshoots, not undershoots. Because after the second star, it is pointing
below the horizon. So not sure what is going on.

I do hear some whining noise mid slewing, but it is not really high, and I don't see the mount stopping.
But to rule out the different microstepping causing a stall, do the above changes, and try again.

Test and see if things are better.


Drew 🔭📷🚴‍♂️
 

On Fri, Jun 11, 2021 at 11:55 AM, Khalid Baheyeldin wrote:
But to rule out the different microstepping causing a stall, do the above changes, and try again.
You forgot to get him to change the
AXIS1_STEPS_PER_DEGREE    12800.0
to 400 (I think) for both axis if you are going to turn off microstepping


Khalid Baheyeldin
 

On Fri, Jun 11, 2021 at 12:22 PM, Drew 🔭📷🚴‍♂️ wrote:
You forgot to get him to change the
AXIS1_STEPS_PER_DEGREE    12800.0
to 400 (I think) for both axis if you are going to turn off microstepping
No.

This value remains as is, since AXIS1_MICROSTEPS is unchanged from 32.

It is only the mode switching that is disabled, in the odd case that the stall was a mid band resonance when doing 16.


Mircea Radutiu
 

That is exactly what happened to my EQ5 when first testing OnStep (MaxESP3) with it, I reversed the Dec. (N-S) direction in the config file and it worked.


Khalid Baheyeldin
 

On Fri, Jun 11, 2021 at 12:56 PM, Mircea Radutiu wrote:
I reversed the Dec. (N-S) direction in the config file and it worked.
Hemendra,

This is easy to try and see if it fixes the issue.

Change the following, then re-upload OnStep:

#define AXIS2_DRIVER_REVERSE ON


Hemendra Rawat
 

Hi Khalid,

I reversed the AXIS2 (Dec axis) and the mount was pointing in opposite direction of the star. So the DEC axis is not the issue here. I reverted the changes.

As suggested by you, I have set the Vref to 0.588 for both the drivers and enabled PULSE mode, and disabled Microsteps GOTO. There is improvement in the sound that motors make but the mount's pointing accuracy is as bad as it was before.

-Hemendra