jif001

There are some photographs and diagrams on page 2 of this thread showing how JeromeW and I independently wired them up. They should help, but feel free to come back if not. You could start by simply replacing the existing motor (I’m pretty sure it’s just a NEMA 14) and checking that it works as it should, before going for the belt mod. if I remember correctly, the arrowed section on the attached diagram shows the original wiring to the RA motor (you may find your NEMA 14 comes with the colours shown in the top right-hand corner of the diagram, but not necessarily!)

I’m happy to send the STLs for the white ones if you’re still interested, but Jerome’s are good.

If you mean the yellow ones then those pics are not of my mount. @JeromeW is your man for those.

As far as I can tell (you may wish to confirm this) the EQ3Pro and the EQM35Pro are the same mount apart from the ability to convert the 35 into a simple RA tracker (i.e. the Dec motor and head are removable - M stands for ‘modular’). If that’s the case, you might find the attached link useful. See my message dated 11 May 2023.

It sounds like you are making some good refinements. Unfortunately I’m still out of commission due to the ongoing house remodelling and will be for a while, in addition to which there has only been one clear night here in months!

The planetary gearbox will have introduced some backlash - perhaps that’s the issue. Do you use PHD2? If so then you can use PHD2 Log Viewer (a separate downloadable application) to analyse the guiding performance and pinpoint issues - see the example plots on the first page of this thread.

It would work the other way. By using a 1.8° motor you would be doubling the overall speed, so you would need to halve it again through the transmission. That would mean using a 200:1 gearbox. Thanks for your input. You’ve moved the dec axis design on substantially from my original effort. I will eventually get a round to printing your designs, but right now we are in the middle of a house reconfiguration and my 3d printer is covered in dust 😬 I’ve recreated some of the components in Fusion 360 to make them editable and that’s kept me entertained for a while! I did wonder though - where does the nut barrel fit in?

I compared your transmission with mine and the original, see attached image, but can see nothing obviously wrong there. Our numbers match (360 motor turns equating to 1 turn of the axis) and differ from the original numbers (720:1) only because we are using 0.9° motors and different gearing (original Skywatcher is a 1.8° motor, resulting in a factor of 2 difference). The fact that you can’t turn the gearbox by hand doesn’t mean there is anything wrong with it - I can’t turn my focus motor by hand, but it’s fine. I’m just at the wrong end of the gear train to apply enough torque to it. What exactly happens? What’s the target if it’s falling 60° short, e.g. are you aiming for 90° and it stops at 30°?

Thanks for the files. The image looks good apart from the lack of calibration, which is easily corrected with darks and flats, and some aberration that indicates the camera is too close to the flattener/reducer. If you look closely, right in the corners, the stars look very slightly elongated in the direction of the image centre (see the image below, which exaggerates the effect). To fix it, you could 3d-print some thin spacers to put between the camera and the flattener/reducer.

Oh yes please!

Assuming your M42 image is not cropped, there is a small amount of aberration in the left-hand corners but it’s not bad at all and the field looks reasonably flat. The stars are a bit out of focus, which is affecting the image more. I don’t know what’s going on down the left edge - a light leak?

You may not need to go back to the drawing board - you may be able to just alter the wiring to the motor. See the extract below from this article. Also, see here (scroll way down the page) - nicely illustrated.

@JeromeW: I don’t want to divert this thread, but I have a quick question … what do you make of the Askar V? I’m considering buying one, but wondered in particular if it can achieve a flat field across all of its configurations. I know there is a flattener, but do the reducer and the extender flatten the field independently?

You seem to be doing with your EQ5 what I did with my EQM35. Have a look at this thread in the Mounts section - there may be some information you find useful.

Mark As you can see in the picture, I engineered some clearance so that the axis could turn fully without clashing, so maybe you need to make some small adjustments to avoid contact (or, as you say, modify the bracket). I used Fusion 360 for the design. There's a bit of a learning curve to it, but there are hundreds of tutorial videos on YouTube. Fusion 360 is very powerful and is used by professionals. For private use it is fully-functional (apart from some bits you will never need anyway), the only real restriction being that you can only have 10 editable projects at one time. I've never found that to be a problem - you can have as many projects as you want as long as only 10 (any 10) are in edit mode. I started out using Tinkercad (also free), which is easy to learn but creating complex designs is tricky. It might be a good place to start though. Jim

Hey it wasn’t my request but thanks anyway - like I said, any improvement is welcome! 😊👍🏻

The holding torque on the motor is pretty good and if you balance your scope it should not challenge that, so it may work in that respect, but the torque necessary to rotate the shaft might be the limiting factor. Turning a worm is a lot easier.

Interesting stuff! I’m assuming that you have eliminated the original dec worm and gear arrangement and replaced it with the orange pulley in your photograph, to be belt-driven by a NEMA 17 motor (how/where will you mount it?). In principle that seems a great idea because it also eliminates the last major source of dec backlash, between the worm and gear. Theoretically, dec guiding should not be needed when imaging because you are tracking in RA only, from east to west. However, reality means that there will be some alignment error or other factor creating a north/south drift that needs to be managed by dec guiding. The corrections needed are generally quite small, which is why dec backlash can be a real headache - it is often larger than the guide corrections, and so the mount appears unresponsive until the backlash has cleared, by which time you may see drift in your image. I really hope your idea works as it should because it seems as if dec backlash should be minimal or non-existent, and if it is then I’ll be making that mod myself! Even though my dec guiding is pretty good (better than RA) any improvement is a worthwhile improvement I’m not sure how the numbers stack up because I can’t picture the exact configuration - maybe post a sketch or a picture of the assembly - but two things are important in deciding the ratios. Firstly you want the dec axis to be responsive enough to allow your guiding software to control it, but not so sensitive that it overshoots with every guide correction. If you are using PHD2 then there are some parameters you can set that may help optimise that. The second important factor is the maximum slew rate - you want it fast enough to be able to locate your target in a reasonable time. Before I made the belt mod, as a benchmark I measured the time taken by each axis to complete a 360-degree rotation when set at the maximum slew rate (Rate 9 on the hand controller). For each it was 105 seconds. For both axes I matched the rates with the original geared configuration. Jim

Great! Post some info here on how it performs when imaging, in particular how it compares with the original configuration: did you see a worthwhile improvement?

In this message on the first page of this thread, I explain how I arrived at the original gear ratios. For dec it was 1:11 using a 1.8 degree motor. You need a ratio of 1:5.5 with a 0.9 degree motor and you can simply adapt those numbers until you get the desired result. So: original figures: Gear teeth: 12 : 66 : 35 : 70 Simple arithmetic using these numbers (noting that the 66 and 35 are co-axial gears so are 1:1): 70/35 = 2 66/12 = 5.5 5.5 x 2 = 11, hence we achieve the desired ratio of 1:11 However, you need 1:5.5 because you’re using a 0.9 degree motor. So, you can adapt those numbers as you see fit to achieve that. Here’s one possibility: 12 : 33 : 20 : 40 40/20 = 2 33/12 = 2.75 2.75 x 2 = 5.5. so, using this example you could have 12 teeth on the motor pulley, a second pulley with 33 and 20 teeth, and 40 teeth on the worm pulley.

This was an issue I had forgotten about. The black wire may be a ground connection, connecting through the board to the motor chassis via those screw holes (see the attached sketch I made at the time), but it doesn’t seem to be essential to the mount operation. I omitted that connection, and the board, in the final configuration and made the connections as shown in the diagram earlier in this thread. The mount has worked fine without it. The board has no components and is merely a way of connecting the two halves of the circuit. If you want to be sure it’s ok, you could use the black wire as a flying lead and connect it to a convenient piece of metal on the mount.

Yeah - I did the same thing!

I’ve been running two NEMA 17 motors for a number of months* and the controller has been fine, so you shouldn’t have any trouble. * when the weather allowed 🙄 but I think I’m well into double figures with imaging hours since making the mods.

You might find this useful. Not an EQ3-2 but I think the EQM35 is much the same in terms of gearing (and if not, you can work out the gearing you need from this)

See my comment above regarding the placement of metal washers on the worm carrier bolts to deal with the misalignment. The mount continues to track very well and accurately, much better than it did with the nylon washers (although this is not necessarily a direct result of the bearing mod - see my other comments on that). A clear conclusion though is that any misalignment can be corrected with a bit of effort, and probably better than the original factory setup.