by

Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    gris74
    @gris74_gitlab
    it's look like the server running on wrong ip adress
    image.png
    the last line of the ofm status after a stop and start of the server
    Joel T. Collins
    @jtc42
    Starting on 0.0.0.0 is normal, that basically just means external devices can access it. Really weird error though. If you close eV and open it again does the stream come back?
    Joel T. Collins
    @jtc42
    @profluid it works fine on the pi4. In fact the one on my desk at home is a pi4.
    Vsaggiomo
    @Vsaggiomo

    @gielium_gitlab here in The Netherlands as well, finished building thee 6.0. ordered the bolts nuts and o-ring from Techniekwebshop.nl (except for the M2 bolts which I ordered from rwproducts.nl). Not cheap though....

    P.S it's on a pi4, no problem there.

    Chris Gielen
    @gielium_gitlab
    @Vsaggiomo ah good :) i'le remember techniekwebshop, i indeed saw the rwproducts.nl very overpriced
    @Vsaggiomo What is your usecase for the microscopes?
    Vsaggiomo
    @Vsaggiomo
    @gielium_gitlab mostly cells and microfluidcs (university). I've put a cheap 40x there, but the working distance is 0.17, need to get a better objective
    Richard Bowman
    @rwb27_gitlab
    very nice :)
    By the way, I've just released v6.1.2 of the hardware design with a few bug fixes (the tools print properly, the motor board fits properly, and the instructions are improved in various places). Raising issues for 404s in the instructions is always helpful, though :)
    Chris Gielen
    @gielium_gitlab
    @rwb27_gitlab haha so i have to restart printing the parts ?:)
    Richard Bowman
    @rwb27_gitlab
    no, the only ones that have changed AFAIK are the motor board case (only applies if you have a Sangaboard 3 PCB) and the actuator assembly tools. The tools haven't changed - but they are now properly aligned within the STL so they all lie flat on the print bed.
    It would be nice to have per-part revision numbers so you don't re-print stuff that hasn't changed - but I have no idea how you'd manage that...
    Gavin S
    @s7726
    If I want to start with the Basic version, but I want the flexibility to go with the higher power version. Can I/Should I print the beamsplitter-compatible body?
    8 replies
    Gavin S
    @s7726
    @rwb27_gitlab have the script that does the 'release' add the release tag to the filenames, but only add it for the last tag it changed in, a little git-fu should be able to spit files that changed between the last tag and the current.
    2 replies
    Julian Stirling
    @julianstirling_gitlab
    The other option is in the CI when we build the docs we use the hashes of the files to work out what changed. We ean then add a note of the last revision each file changed.
    22 replies
    gris74
    @gris74_gitlab
    @jtc42 Thx for your answer, indeed it come back normal after a 3th complet rebuild of the Raspian openflexure, the first too installation with the probleme was bacause I had get version 2.0.beta1 of the server
    gris74
    @gris74_gitlab
    at the end i dont understand ? no more image after a wild of use tooday, and after a check again I have version 2.0.0-beta.6-Lite without doing any update and same problem of losing image on web an eV windows version
    Gavin S
    @s7726
    @julianstirling_gitlab any chance you have some advice on choosing a base?
    Going to start with the basic camera setup, but I'd like to not have to reprint the base to upgrade the optics etc. later. Can I just print the 'better' base now and use it or will it not work with the lower end camera/optics?
    James Haley
    @jhaleyaero_gitlab
    well update on the stage not being detected, after giving up and doing nothing for a few days, i plugged back in microscope / stage, and it is detected now... not sure why, focus seems to work too.
    only thing that has changed that I can think of, is i shuffled the USB port connections around
    orgemd
    @orgemd
    I am just wrapping up my first microscope build. I am going as simple as possible to start - so just RPi 4B with Picam v2 - no motors or additional driver boards, no additional optics, nothing beyond simple white LED illumination... From Pi interface, eV will connect local and show image, but then will crash leaving just the image permanently on the screen and a white window around it. I can close the eV app, but the microscope image stays until the Pi is rebooted. Both Windows 10 and Linux systems often get network errors when I try to use eV from them - though both ssh without issues. Sometimes either will connect and show the image for 1-2 seconds. The image then goes white. The image I am seeing is promising, but I can't get consistent, controlled display. Any guidance?
    Julian Stirling
    @julianstirling_gitlab
    @s7726 You might want to look at the STL selector that Kaspar is developing. You can then see which options use each stl. The website is still in development but I think it could help https://ofm-stl-selector.netlify.app
    gris74
    @gris74_gitlab
    @orgemd same issue here and I solve it by redo the SD 3th time, but now I have issue with the stage not reconnize and network error when arduino is plug
    orgemd
    @orgemd
    OK. Thanks @gris74_gitlab . I will download a fresh copy and try reflashing the SD and see if that helps. I assumed the SD was good since Raspbian looks to be running fine, but definitely worth a shot if you had similar issues and saw some improvement...
    gris74
    @gris74_gitlab
    I try again this trick of fresh install from Rasbian with the stage inplace but no luck I still have ERR_CONNECTION_REFUSED on browser or Network Error on eV software :o/
    any advices ?
    I follow the instruction assembly on the motor section and wired as picture on an arduino uno flash with Sangaboard v0.3
    If I unplug the arduino every thing go well..
    10 replies
    gris74
    @gris74_gitlab
    And if I hotplug back the arduino (without reboot) the microscope work fine but Stage:
    No stage connected
    orgemd
    @orgemd
    I flashed the SD with the 2020-05-20 Raspbian build. The connectivity errors seem to be gone for me with that version. So far, no connection refused from direct, linux, or windows. The image in the stream still disappears in remote connections, and becomes locked on the screen in direct connections. Captures work - you just don't know what you are capturing until it saves the image.
    Gavin S
    @s7726
    @julianstirling_gitlab That's a neat tool. Doesn't quite answer the question (or maybe it does). From fiddling with options on the tool, it looks like I can use the one base for non-reflection illumination; but that if I ever want reflection I need a new base. I was hoping I could make both work with only ever having to print one base, the 'big' long-print part.
    3 replies
    Julian Stirling
    @julianstirling_gitlab
    The refection main body will work for transmission. It is just a little less strong and requires a bucket-base as the hole for the reflection optics means you cannot connect a back foot. Was helpful for you to look at the tool and to find out what is still unclear
    orgemd
    @orgemd
    I take it back - there are still network errors. It is just luck of the draw on any microscope reboot whether or not a connection can be made. I got lucky on the first test boot after configuration and at least network connectivity looked good (though the stream was still inconsistent). With subsequent boots, it is more likely than not that connectivity will not work.
    8 replies
    Gavin S
    @s7726
    @julianstirling_gitlab That's awesome! Thanks... I can probably come up with some more feedback if you're open.
    I'm seeing something like a tech-tree. I think that might help visualize the available upgrade paths, if not for actual consumption, to help drive the include/exclude options in the tool.
    Julian Stirling
    @julianstirling_gitlab
    https://gitlab.com/openflexure/ofm-stl-selector If you could make an issue on the GitLab page then we will have a permanent record of the suggestion. A sketch of the idea would be useful. We are going to be doing a big push towards improving the docs trying to automate bills of materials with a tool we are writing GitBuilding (gitbuilding.io).
    It would be interesting to think about what data would be needed to visualise upgrade paths.
    aboomer
    @aboomer
    Firstly, thank you for a great project! I'd like to make a microscope for my children (eldest is 7 years) and I would appreciate some advice (from anyone will to give it). I started with the basic optics build but in hindsight I see two issues, firstly it has too much magnification and second it doesn't support reflective illumination. I think my children will initially be more interested in looking at parts of small bugs etc. so less magnification and reflective illumination would be easier for them to grasp. As I understand it, this means I should be going down the proper rms objective route. As these lenses are more pricey and have longer delivery times, I'd appreciate any advice on amount of magnification that would be recommended and ideally any suggestions for specific objectives too? Overall image quality is less of a concern for me, anything on par with the quality of the Raspberry pi camera v2 setup (that I currently have) would be fine.
    4 replies
    Julian Stirling
    @julianstirling_gitlab
    It is much harder to get reflective illumination to work well. I am not sure whether it is easier to understand. It should be possible to build one using a pi camera lens that has a lower magnification by adjusting where the lens is held. @rwb27_gitlab do we have a way of generating other magnifications for the pi lens?
    If you do go down the RMS route there are some example images I have taken with the x4 and x10 objectives here: https://gitlab.com/openflexure/gallery-wiki/-/wikis/Field-testing-the-OpenFlexure-in-the-Panama-rainforest
    Oleg Arefyev
    @tarkus.su_gitlab
    Hello, tell me how much the image will deteriorate if you do not install a lens with a diameter of 12.7 focus 50mm when using a RMS x40/160 objectives? Thanks!
    1 reply
    aboomer
    @aboomer
    @julianstirling_gitlab, thank you for your input. Looking at your example images I reckon that x4 would probably be ideal for the kinds of things my children would want to look at. My thinking about reflective illumination being easier for them to understand was from the point of view of recognising an object better with light shining on it (rather than through it). In terms of difficulty in operating the microscope with reflective illumination...I would expect to be setting all this up for them when they have a subject to look at.
    5 replies
    (sorry hit enter prematurely). Was going to ask, are your example images all done with transmission illumination? Is the bee leg simply lit with ambient light? If so, perhaps with less magnification, the need for reflective illumination is maybe less important?
    orgemd
    @orgemd
    What are most people doing to add motor control? Are either of the Sangaboard versions available for purchase someplace with all of the electronic components already integrated? Or are people just buying Arduinos and using ULN2003 drivers? I like the idea of an integrated board, but would not want to try to solder all the components on myself...
    orgemd
    @orgemd
    Another question - is anyone using 100x objectives, or are most using 40x? I see in the build docs that it says up to 100x. In my experience, however, 100x will not work without immersion oil. I can't see how that would work with the inverted design. If people are using 100x, is there some trick to use immersion oil, or does this design have some other way to deal with the refraction problems?
    5 replies
    Julian Stirling
    @julianstirling_gitlab
    @orgemd For motor boards I think everyone is doing the Arduino and ULN2003 drivers. V0.2 of the sangaboard can be had soldered if you get the board made. v0.3 and v0.4 are surface mount and require assembly at a board house. We really want to start selling v0.4 soon, but we need to do EMC testing, which has been delayedt during lockdown. Also I think there was a problem with the first round of v0.4 boards that needs fixing.
    Julian Stirling
    @julianstirling_gitlab
    @/all Happy to announce that a preprint of our paper on our collaboration is now online at AfricArXiv https://twitter.com/AfricArxiv/status/1266232895452020736
    orgemd
    @orgemd
    @julianstirling_gitlab - OK - thanks for that, I will get a standard nano, the recommended steppers, and the ULN2003 boards for now. The good news is that those are available pretty much everywhere.
    Richard Bowman
    @rwb27_gitlab
    For those who were asking about low magnification, Al Edwards at Reading University worked with a project student to produce this - showing the field of view you get when using different lengths of spacer tube between the pi camera and pi camera lens.
    image.png
    Julian Stirling
    @julianstirling_gitlab
    @/all I am excited to announce that we now have a forum provided by Discourse: https://openflexure.discourse.group/ . Hopefully this will be simpler to follow that some of the threaded discussions on here :D