Topics

Newbie Dobsonian Project


Ken
 

Hello Group,

Pardon my engineering ignorance but what formula do you use to calculate the teeth needed on the motor pulleys and corresponding alt-az gears to match a certain stepper motor with planetary drive? I am thinking driving the bearings by using gears (versus belts) would be the way I would try to convert my dob to an OnStep goto Dob. Tony Star and Mihai Gavrilciuc's designs seem the best way to achieve this goal. I have read the wiki, showcases, and the spreadsheet and perhaps I am not fully understanding the output value meanings. I have seen some use a 51:1 or 27:1 planetary Nema steppers with a pulley attached to them that drive a fastened belt on the dobsonian ALT-Az bearings. Is the GR1 the planetary gear ratio and then GR2 being the ratio of the pulley attached to the motor and the larger bearing with a belt glued/epox to it?

Ken


Khalid Baheyeldin
 

On Tue, Dec 1, 2020 at 01:18 PM, Ken wrote:
I have seen some use a 51:1 or 27:1 planetary Nema steppers with a pulley attached to them that drive a fastened belt on the dobsonian ALT-Az bearings. Is the GR1 the planetary gear ratio and then GR2 being the ratio of the pulley attached to the motor and the larger bearing with a belt glued/epox to it?
GR1 is usually the gear reduction from the motor shaft, and GR2 is any reduction after that.

So on an equatorial mount, GR1 will be the pulley ratio from the motor shaft to the worm gear, and GR2 will be the the worm wheel number of teeth (e.g. 144:1 to an EQ5 mount).

For Dobsonians, it does not matter which reduction goes where since there is no worm gear, and there is no periodic error on Alt-Az mounts. So you can swap GR1 and GR2.

The idea though is to have enough gear reduction to get you good resolution, and increase torque.
So a geared planetary with 50:1 or 100:1 is a good start. It allows a smaller pulley for the 2nd stage.

Here is an example of a 100:1 geared motor, and a 4:1 pulley reduction (12T to 48T, or 15T to 60T):

200, 64, 4, 100 = 14222.222 steps per degree.
This gives you 0.25"/step (with an estimate of up to 0.87"/step), which is great for visual.

The mechanical part is always a challenge, and once you figure it out, within the constraints above, the rest is a known quantity.


Ken
 

Alright so using my base as the gear (via glued/epoxied belt) with 605 teeth and a pinion with 24 teeth, HDT 3mm pitch/ 9mm wide and a 51:1 nema 17 stepper motor (and similar for Alt bearing) it comes out to:


I am trying to keep it simple and follow what Mihai's plan did here: https://onstep.groups.io/g/main/topic/31872764#10214


Khalid Baheyeldin
 

On Wed, Dec 2, 2020 at 01:29 PM, Ken wrote:
Alright so using my base as the gear (via glued/epoxied belt) with 605 teeth and a pinion with 24 teeth, HDT 3mm pitch/ 9mm wide and a 51:1 nema 17 stepper motor (and similar for Alt bearing) it comes out to:
Looks good.
Only change is from 1/64 microsteps to 1/32, since you don't need as fine resolution, and it would be better step accuracy, and faster slewing.

Look in the Showcase page for Raymon Collecutt's 20" Dobsonian. He placed the motor on a hinge so he can disengage the motor during transportation, and (not clear from his pictures) there is a mechanism to engage it.


Dave Schwartz
 

That's very fine tracking for a DOB with homemade pulleys and belts. Its trading off slewing speed for resolution but I'm not sure you'll be able to take advantage of that.

Although STEPS_PER_WORMROT is calculated as some crazy-high number, set it to zero when you create the Config.h file because you can't use PEC with an ALT/AZ DOB.

On 2020-12-02 1:29 p.m., Ken wrote:
Alright so using my base as the gear (via glued/epoxied belt) with 605 teeth and a pinion with 24 teeth, HDT 3mm pitch/ 9mm wide and a 51:1 nema 17 stepper motor (and similar for Alt bearing) it comes out to:


I am trying to keep it simple and follow what Mihai's plan did here: https://onstep.groups.io/g/main/topic/31872764#10214 <https://onstep.groups.io/g/main/topic/31872764#10214>;


Ken
 

Thanks for the feedback. I began searching for parts and found that GT2 belts with a 2mm pitch in the length I need is widely available compared to HDT / 3mm. I recalculated using 908 teeth for the Az gear and a 30 tooth gear pinion. Do you all see any disadvantages of using a 2mm pitch vs 3mm if the final stage is ~ the same? I would rather have a little better tracking than slewing speed (I think) but than again I can always replace the 30s with 40s or more.


Khalid Baheyeldin
 

On Mon, Dec 7, 2020 at 01:22 PM, Ken wrote:
Do you all see any disadvantages of using a 2mm pitch vs 3mm if the final stage is ~ the same?
The important thing is that the teeth type on both pulleys and the belt must match.
They have to all be GT2 or MXL or whatever. Do not mix the types.
We had a long thread on an large Astro-Physics that was not tracking right because of mixing pulley types.
That caused the actual ratio to be different from the one in the spreadsheet.

As for pulley size, even a 12T pulley on the motor shaft can drive a huge mount, so don't fret about that.

One other thing to think about is portability: if you plan to move your scope around in a car, how will the
mechanics between the motor and last stage stand up to being bumped around? You don't want the
belt to slip or something ...


Ken
 

I received the motors, lazy susan and some misc elec items. I submitted an order for the FYSETC S6 board and now am calculating what drivers to choose. I get that they must match the motor voltage and current. Motor rated current: is 1.68A, Phase Resistance: 1.65ohms, Voltage: 2.8V, Inductance: 2.8mH ± 20%(1KHz). With these numbers I think I am limited to a few drivers, the TMC5160, TMC2209, S109, and DRV8825 . TMC5160s min amps were limit 2.0A, max 3.0 on the wiki so am I correct in thinking these would work with the motors even though the voltage is way above range? TMC2130 almost looks good but I am not sure if these would be able to drive the motor due to the max being 1.5 for these drivers. The vendor spec sheet says max is 2.5A, 1.2 RMS (https://wiki.fysetc.com/TMC2130/). 


Khalid Baheyeldin
 

With 1.68A motors, your range will be:

Min 1.68 x 1.41 x 0.4 = 0.95A
Max 1.68 x 1.41 x 0.7 = 1.66A

That means only the ones you mentioned are rated for the possible value you will end up with.
The DRV8825 are the worst, and should be avoided due to vibration.
The TMC2209 works, but not sure if it can do mode switching on the S6 or not. It also has a limited range of microstepping.
The TMC5160 are the best overall. I am not sure what that minimum of 2A is from. I am using them at 0.8A for slewing and 0.67A for tracking, and they work well on my S6. Go with those.
I am using the BigTree Tech V1.2 variants. Did not need to solder any wires or cut any pins.


Jesse Lichtenberg
 

If you picked up the motors from Stepper Online and they are the economy 51:1 models, take a look at the specs sheets for the accurate gearing ratio. It's not a straight 51:1, but more like 50.895:1


Ken
 

On Wed, Dec 16, 2020 at 05:16 PM, Jesse Lichtenberg wrote:
50.895
All my parts came in and I have soldered everything I needed to, flashed the S6 + Wifi. Connected a GPS as well. I can connect to it via web and am in the process of building out wiring to the steppers via Ethernet, combining 2 channels for each of the 4 wires.


Ken
 

I have pretty much everything connected and flashed. The gps does not seem to be working and the motors do not turn at all. WiFi works though. I am getting a nv EPROM error so I am going to try and enable debugging and watch the loader via the serial monitor to further troubleshoot. 


Ken
 

Alright here is the output with OnStep Loading without the GPS or WiFi connected. Error MSG is in bold and after reading some threads it seems like its related to PEC? This is Alt/Azm so I isn't PEC ignored? It also says the stepper drivers are enabled and then disabled.

MSG: OnStep 5.1k
MSG: MCU =  STM32F446, Pinmap = FYSETC S6
MSG: Init HAL
MSG: Init serial
MSG: Init pins
MSG: Init TLS
MSG: Start NV 2048 Bytes
MSG: Init NV Axis1 defaults
MSG: Init NV Axis2 defaults
MSG: Init NV Axis3 defaults
MSG: Init NV Axis4 defaults
MSG: Init NV Axis5 defaults
MSG: Read NV settings
ERR, initReadNvValues(): pecBufferSize exceeds available NV, PEC disabled
MSG: Init startup settings
MSG: Init library/catalogs
MSG: Init guiding
MSG: Init weather
MSG: Init sidereal timer
MSG: Init motor timers
MSG: Axis1/2 stepper drivers enabled
MSG: Setting up Axis1/2 TMC stepper drivers
MSG: Axis1/2 stepper drivers disabled
MSG: Serial buffer flush
MSG: OnStep is ready

config.h

#define PEC_SENSE                     OFF //    OFF, ON*, n, sense digital OR n=0 to 1023 (0 to 3.3V or 5V) analog threshold. Option
#define PEC_SENSE_STATE              HIGH //   HIGH, Senses the PEC signal rising edge or use LOW for falling edge.           Adjust
                                          //         Ignored in ALTAZM mode.


Howard Dutton
 
Edited

On Thu, Jan 14, 2021 at 06:30 AM, Ken wrote:
This is Alt/Azm so I isn't PEC ignored?
No.

Probably you really should be using the beta branch, not the master.  The master isn't well tested at the moment.
Probably AXIS1_STEPS_PER_WORMROT should be 0 (the default.)
Definitely just saying you don't have a PEC sensor doesn't imply PEC is disabled, as it can work without a PEC sensor.
OnStep (attempts to) disable the axis1/2 stepper drivers at startup if possible.  That's normal.

Really the log shows nothing critically bad happened at boot; there was an unexpected PEC buffer size, which it fixed, and that's it.


Howard Dutton
 

I would focus on getting the TMC drivers working.

Make sure the S6 board shunts (connecting the SPI signals to the driver pins) were positioned correctly (follow the S6 Wiki.)
Make sure they are the correct TMC drivers mfg/model and any suggested modifications were carried out per the Wiki instructions:
SilentStepStick TMC2130 or TMC5160 (read about these drivers before buying!)


Ken
 

Thanks, Howard. I changed to beta and set AXIS1_STEPS_PER_WORMROT to 0. The error went away and the GPS started to work. Now trying to figure out why the motors do not turn.


Howard Dutton
 

On Thu, Jan 14, 2021 at 06:50 AM, Howard Dutton wrote:
On Thu, Jan 14, 2021 at 06:30 AM, Ken wrote:
This is Alt/Azm so I isn't PEC ignored?
No.
Hint... the master branch allows switching between Alt/Azm and equatorial modes at runtime.


Ken
 

I re-checked the S6 Wiki and manufacture wiki. I am using TMC5160 drivers from BigTreeTech v1.2. Please note I did NOT cut off the CLK pin as suggested in the OnStep drivers wiki because I have read on this board that this wasn't needed with the FYSETC S6 controller. Maybe because CLK is pulled to GND on the 1.2 version. My jumpers look like this:

I am using the Android APP and it allows me to unpark but actions that should initiate the motors to move do nothing. The motors are attached to the board but sitting on my workbench so there is no telescope mount involved at this point.


Howard Dutton
 

On Thu, Jan 14, 2021 at 10:07 AM, Ken wrote:
Please note I did NOT cut off the CLK pin as suggested in the OnStep drivers wiki because I have read on this board that this wasn't needed with the FYSETC S6 controller. Maybe because CLK is pulled to GND on the 1.2 version.
That may very well be the case, the Wiki advice is generic.


Khalid Baheyeldin
 

Did you set the location correctly? Time? Date? UTC offset?

If you don't, then OnStep may think that it is pointing beyond the limits allowed (e.g. below the horizon), and refuse to move the motors.