3D mount file has error

Cognisant1 commented 3 years ago

UsingCQ-editor, the Mount does not work properly. Before the circle points are added, the workplane needs to state (centerOption="CenterOfMass") otherwise the motor Mount points are on the edge of the board rather than the centre.

Peque commented 3 years ago

It may be due to changes in the latest 2.1. Before, I had seen:

cadquery/cadquery/utils.py:21: FutureWarning: Default walue of centerOption will change in the next relase to ProjectedOrigin

Don't know which was the previous default centerOption.

Peque commented 3 years ago

Yeah, it was changed in 2.1:

https://github.com/CadQuery/cadquery/blob/master/changes.md

@Cognisant1 Want to create a PR with the changes? :wink:

Cognisant1 commented 3 years ago

Hi. I’ve got the version of the file that works but this is my first action in GitHub so I don’t know how to post it. Any help would be great!

Sent from my iPhone

On 15 May 2021, at 15:53, Miguel Sánchez de León Peque @.***> wrote:

Yeah, it was changed in 2.1:

https://github.com/CadQuery/cadquery/blob/master/changes.md

@Cognisant1 Want to create a PR with the changes? 😉

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Peque commented 3 years ago

@Cognisant1 Maybe this is a good starting point:

https://docs.github.com/en/github/collaborating-with-issues-and-pull-requests/proposing-changes-to-your-work-with-pull-requests

:blush:

Cognisant1 commented 3 years ago

That was very messy - sorry I've now added the modified mount.py file (and the mount.stl and rim.stl files, in case they'd be useful) in to this zip folder.i hope that's better practice! Mount_OK.zip

Peque commented 3 years ago

@Cognisant1 I'd say, in general, it is better to share your code changes by either:

Sharing the code diff in a comment, if it is just a couple of lines (you can execute git diff and copy-paste the output)
Pushing your changes to a branch in your fork and sharing that branch or, even better, opening a pull-request with the changes! :wink:

Do you want to try with the pull-request?

Cognisant1 commented 3 years ago

Yes that would be good, but I’m still not sure how to open a pull request. Do you think the stl files are worth adding??

Sent from my iPhone

On 16 May 2021, at 11:54, Miguel Sánchez de León Peque @.***> wrote:

@Cognisant1 I'd say, in general, it is better to share your code changes by either:

Sharing the code diff in a comment, if it is just a couple of lines (you can execute git diff and copy-paste the output) Pushing your changes to a branch in your fork and sharing that branch or, even better, opening a pull-request with the changes! 😉 Do you want to try with the pull-request?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Peque commented 3 years ago

@Cognisant1 I'd rather not add the STL files. Instead, I would prefer to pin the CadQuery version in the environment, to make sure 3D designs are reproducible and do not break with future CadQuery updates.

Cognisant1 commented 3 years ago

Hi Miquel

Is there a reason why you didn't 3D print the gear onto the rim. It would avoid centring errors.

Sent from my iPhone

On 16 May 2021, at 13:33, Miguel Sánchez de León Peque @.***> wrote:

@Cognisant1 I'd rather not add the STL files. Instead, I would prefer to pin the CadQuery version in the environment, to make sure 3D designs are reproducible and do not break with future CadQuery updates.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Peque commented 3 years ago

@Cognisant1 Indeed!

The reason was that we got some good second hand motors really cheap from Aliexpress (at the time they were 8 or 15 € each!), so we designed the locomotion system around them. The motors came with a 0.3 modulo gear soldered to the shaft and we didn't have access to a precise-enough 3D printer, so we couldn't print wheel gears that were compatible with the motor...

Nowadays, with resin 3D printing you could print even a 0.3 modulo gear I think, so it should be easier! :blush:

Please, share any modifications you make to the locomotion system! We'd love to know about different designs that do not require the currently hard-to-find-second-hand (or otherwise too expensive) Faulhaber motors. :wink:

Cognisant1 commented 3 years ago

Hi Miguel

I've already designed and printed a rim with integrated 60 tooth 0.3 modulo gear. But the alternative motor is a real problem. The Faulhaber encoder is a real plus. The ordinary N20 motor encoders have very poor resolution, as you know. The futura mouse encoder system is much better but the wheel based magnetic disc is expensive to obtain from China (at least in small quantities).

I would be happy to work on this a bit...I'm 81.. so plenty of time😄.

Sent from my iPhone

On 18 May 2021, at 18:05, Miguel Sánchez de León Peque @.***> wrote:

@Cognisant1 Indeed!

The reason was that we got some good second hand motors really cheap from Aliexpress (at the time they were 8 or 15 € each!), so we designed the locomotion system around them. The motors came with a 0.3 modulo gear soldered to the shaft and we didn't have access to a precise-enough 3D printer, so we couldn't print wheel gears that were compatible with the motor...

Nowadays, with resin 3D printing you could print even a 0.3 modulo gear I think, so it should be easier! 😊

Please, share any modifications you make to the locomotion system! We'd love to know about different designs that do not require the currently hard-to-find-second-hand (or otherwise too expensive) Faulhaber motors. 😉

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Peque commented 3 years ago

Yeah, the Faulhaber encoders are great. :stuck_out_tongue_winking_eye:

There are a couple of references in Alibaba (https://spanish.alibaba.com//trade/search?fsb=y&IndexArea=product_en&CatId=&SearchText=1524B009SR), but I think I once contacted them and they did not have any available in reality... Maybe you have a higher chance with the more expensive seller, if it fits within your budget.

Anyways, software can do magic so, even if it will be harder, you can achieve a good performance with much worse encoders. Also, the Futura mouse is a great reference, but you may find this project interesting (with pretty easy-to-find encoders):

https://github.com/ukmars/ukmarsbot/wiki

:wink:

Cognisant1 commented 3 years ago

Yes. I've already built a ukmarsbot. It's good but not very fast🙁. Peter Harrison is a great software and electronic genius. He managed to double the pololu encoder count using a quad XOR gate and his profilers give a very smooth ride. The Nano is slow and a bit limited in memory. I liked your blue pill and the gyro. Maybe I just need to save up for the Faulhaber motors 😣

Sent from my iPhone

On 18 May 2021, at 20:42, Miguel Sánchez de León Peque @.***> wrote:

Yeah, the Faulhaber encoders are great. 😜

There are a couple of references in Alibaba (https://spanish.alibaba.com//trade/search?fsb=y&IndexArea=product_en&CatId=&SearchText=1524B009SR), but I think I once contacted them and they did not have any available in reality... Maybe you have a higher chance with the more expensive seller, if it fits within your budget.

Anyways, software can do magic so, even if it will be harder, you can achieve a good performance with much worse encoders. Also, the Futura mouse is a great reference, but you may find this project interesting (with pretty easy-to-find encoders):

https://github.com/ukmars/ukmarsbot/wiki

😉

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Peque commented 3 years ago

@Cognisant1 Ah, not your first micromouse then! :smile:

Yeah, the Bluepill is enough for not-very-complex control algorithms. I also started some work with a new Bluepill-like F4 board (https://github.com/Peque/bulebule/commit/2b3799c71168a3d3fbc35a4173a083a511ced357), but I would not consider it as stable... Only recommended if you want to have some fun with hardware design and testing! :stuck_out_tongue_winking_eye:

Here is a fork of the Bulebule project with a different locomotion system:

https://github.com/hmightypirate/AllMyCircuits/wiki/Buriburi-(Micromouse)

I think the motors are a little too weak, but maybe you can take it as an inspiration for your own design. A very common design is with the motors between the wheels and the encoder at the front/back wheel's shaft. Note that those motors are quadcopter's motors, very cheap and with high power. Encoders are cheap too, but require a little bit of work to design and mount.

Here I did some research for the new board (and with new encoders):

https://github.com/Bulebots/bulebule/issues/428

The as5047p looked like a great contender. :wink:

Cognisant1 commented 3 years ago

Wow, Miguel

What a feast of wonderful information!! Thank you so much😁

I'll have fun digesting all this. 😊

Sent from my iPhone

On 19 May 2021, at 11:43, Miguel Sánchez de León Peque @.***> wrote:

@Cognisant1 Ah, not your first micromouse then! 😄

Yeah, the Bluepill is enough for not-very-complex control algorithms. I also started some work with a new Bluepill-like F4 board @.***), but I would not consider it as stable... Only recommended if you want to have some fun with hardware design and testing! 😜

Here is a fork of the Bulebule project with a different locomotion system:

https://github.com/hmightypirate/AllMyCircuits/wiki/Buriburi-(Micromouse)

I think the motors are a little too weak, but maybe you can take it as an inspiration for your own design. A very common design is with the motors between the wheels and the encoder at the front/back wheel's shaft. Note that those motors are quadcopter's motors, very cheap and with high power. Encoders are cheap too, but require a little bit of work to design and mount.

Here I did some research for the new board (and with new encoders):

428

The as5047p looked like a great contender. 😉

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Cognisant1 commented 3 years ago

Hi Miguel

Really sorry to bother you again, but it seems impossible to source the 5v and 9v boost boards to fit the PCB layout. All the boards now seem to have the ground as the middle pin whereas yours has the ground at one end. Can you please indicate where these boards came from or send me the Eagle (or equivalent files) so that I can slightly modify the PCB layout? Modifying the gerber files themselves is really difficult!

You’ve been so kind in your help so far that I hope I can ask for this extra assistance!

Best wishes

Adrian

On 19 May 2021, at 11:43, Miguel Sánchez de León Peque @.***> wrote: @Cognisant1 Ah, not your first micromouse then! 😄

Yeah, the Bluepill is enough for not-very-complex control algorithms. I also started some work with a new Bluepill-like F4 board @.***), but I would not consider it as stable... Only recommended if you want to have some fun with hardware design and testing! 😜

Here is a fork of the Bulebule project with a different locomotion system:

https://github.com/hmightypirate/AllMyCircuits/wiki/Buriburi-(Micromouse)

I think the motors are a little too weak, but maybe you can take it as an inspiration for your own design. A very common design is with the motors between the wheels and the encoder at the front/back wheel's shaft. Note that those motors are quadcopter's motors, very cheap and with high power. Encoders are cheap too, but require a little bit of work to design and mount.

Here I did some research for the new board (and with new encoders):

428

The as5047p looked like a great contender. 😉

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Peque commented 3 years ago

@Cognisant1 No problem! :blush:

9 V converter (pack of 10... we donated the spare parts to a local fablab to avoid waste, or you can try to find single pieces :wink:)
5 V converter
3.3 V converter (again, pack of 5...)
Motor driver
Gyro board
Bulebule schematics and PCB design, feel free to fork and make any modifications you want! :blush:

Feel free to add this info to the project documentation, I would happily accept a merge request. I didn't want to sponsor Aliexpress but... it may be useful for people trying to build the robot to at least have one option to buy the parts.

Cognisant1 commented 3 years ago

That’s brilliant, Miguel. Thank you so much.

Adrian

Sent from my iPhone

On 11 Jun 2021, at 11:08, Miguel Sánchez de León Peque @.***> wrote:

@Cognisant1 No problem! 😊

9 V converter (pack of 10... we donated the spare parts to a local "fablab" to avoid waste, or you can try to find single pieces 😉) 5 V converter 3.3 V converter (again, pack of 5...) Motor driver Gyro board Bulebule schematics and PCB design, feel free to fork and make any modifications you want! 😊 Feel free to add this info to the project documentation. I didn't want to sponsor Aliexpress but... it may be useful for people trying to build the robot to at least have one option to buy the parts.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Cognisant1 commented 3 years ago

Hi Miguel

I’ve decided to try to get the faulhaber motors even though they are so expensive. However I have had this reply from one of the suppliers: “ Unfortunately, Faulhaber do not offer the motor encoder with the pinion you require”

Can you help at all?

Best wishes

Adrian

Sent from my iPhone

Peque commented 3 years ago

@Cognisant1 If you are thinking about buying brand new motors, maybe you should consider the 1717 instead (I think they are about the same price). Those are the ones we wanted to use for our Meiga design. They have a couple of advantages:

They allow for a slimmer robot design
They have a higher no-load speed (14 000)
Maybe others (don't remember :joy:)

Think whether you want the 3 V version or the 6 V version... I think @micromouseonline used the 3 V version. You'll need less battery cells in series to drive the motor but you'll have higher current peaks. If you have a look at 2018 All Japan contestants' hardware you'll find many Faulhaber with both 3 V and 6 V versions so... Both are fine I guess. Note the extended use of the 1717 and no 1524.

You can buy the pinion/gears from another shop (this one is an example here in Spain). Here again you need to decide you preferred modulo. Bulebule uses 0.3, but @micromouseonline uses 0.5 or 0.6 (if I remember correctly) and got 2nd place in 2017 All Japan! So not that you need lower modulo to run faster. :stuck_out_tongue_winking_eye:

Also, if you are doing a "tetra" design, think about the whole design of the locomotion system (tires, gears, pinion, PCB distance to floor...). It all goes together. We documented a quick guide about tetra locomotion systems if you want to have a look. You can easily visualize your own configuration with Python.

Last but not least, make sure you are asking your supplier for the motor plus the encoder! (a Faulhaber motor by default comes with no encoder). If you forgot that... the new price will probably double. :sweat_smile:

If you would rather make your own encoders, then I think Faulhaber motors are not worth it. Just use common quadcopter motors instead, which are much cheaper. :wink:

micromouseonline commented 3 years ago

If your goal is top performance use the 1717 motors with encoders but you may be shocked at the price. I use 3V parts and drive them from a 2S LiPo battery with a maximum drive voltage of 6.6Volts.

My gears are 0.5mod because they are cheap, easy to find and more tolerant of positional errors.

On 02/07/2021 17:26, Miguel Sánchez de León Peque wrote:

@Cognisant1 https://github.com/Cognisant1 If you are thinking about buying brand new motors, maybe you should consider the 1717 instead (I think they are about the same price). Those are the ones we wanted to use for our Meiga design https://upverter.com/design/bulebots/meiga/. They have a couple of advantages:

They allow for a slimmer robot design

They have a higher no-load speed (14 000)

Maybe others (don't remember 😂)

Think whether you want the 3 V version or the 6 V version... I think @micromouseonline https://github.com/micromouseonline used the 3 V version. You'll need less battery cells in series to drive the motor but you'll have higher current peaks. If you have a look at 2018 All Japan contestants' hardware http://www.ntf.or.jp/mouse/micromouse2018/recode2018_181215.pdf you'll find many Faulhaber with both 3 V and 6 V versions so... Both are fine I guess. Note the extended use of the 1717 and no 1524.

You can buy the pinion/gears from another shop (this one is an example here in Spain http://www.mootio-components.com/_refe_008472.html). Here again you need to decide you preferred modulo. Bulebule uses 0.3, but @micromouseonline https://github.com/micromouseonline uses 0.5 or 0.6 (if I remember correctly) and got 2nd place in 2017 All Japan! So not that you need lower modulo to run faster. 😜

Also, if you are doing a "tetra" design, think about the whole design of the locomotion system (tires, gears, pinion, PCB distance to floor...). It all goes together. We documented a quick guide about tetra locomotion systems https://bulebots.readthedocs.io/en/latest/tetra.html if you want to have a look. You can easily visualize your own configuration with Python https://github.com/Bulebots/bulebots/blob/master/docs/source/tetra.py.

Last but not least, make sure you are asking your supplier for the motor plus the encoder! (a Faulhaber motor by default comes with no encoder). If you forgot that... the new price will probably double. 😅

If you would rather make your own encoders, then I think Faulhaber motors are not worth it. Just use common quadcopter motors instead, which are much cheaper. 😉

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Bulebots/bulebule/issues/437#issuecomment-873117309, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAFC3KQUPLADOOUSDQJTWIDTVXSDNANCNFSM445T3EVQ.

-- Peter Harrison

Bulebots / bulebule

3D mount file has error #437

428

428