This is an experimental version of the OFM that uses the BIGTREETECH SKR Pico V1.0 board for driving NEMA14 steppers, uses klipper + fluidd for the UI and easy movement, with crowsnest to stream the view from the microscope. Also testing how the main body behaves after a filament change / pause since these will be printed in places where electricity is unreliable.
It features an alternative setup for the illumination - using a photography diffuser glued to a printed part. It runs a high CRI LED mounted on a smart clothes PCB (which is essentially a resistor, LED that is swapped out for a high CRI Osram LED and two pads that we screw into).
Still some ways to go to get this working with the V3 server software. I’m able to manually tune the webcam, but don’t know too much about photography to get a decent image.
I’m hijacking the board’s ARGB header for the 5V and ground to power the LED. I’ll check if I’m able to get PWM to get LED dimming. I’m not too good at electronics to know how ARGB works.
Need to change up the tray to set the mouting points for the SKR Pico, and to fit a dummy USB-C PCB that’ll allow me to pull 24V for the microscope. Right now it’s going off of a lab power supply. Might need to make the tray a bit wider for everything to fit though.
High CRI does not completely cover the requirements for scientific microscopy. White LEDs usually combine a blue LED with a phosphor that has a peak emission in the yellow. Many LEDs get a high CRI, but still have almost no light in the turquoise/true green region (480-520nm).
The LED specified in the illumination PCBs are Bridgelux Thrive, which have a proper continuous spectrum. I think that they achieve this with multiple/mixed phosphors. We have chosen the 5000K colour temperature version that has a flat intensity across the visible. This will be slightly bluer than a halogen bulb, but the diffuser is likely to scatter the blue more. I have not yet measured the actual spectrum delivered to the sample.
So I have gotten OFM server to work, barring the things that are connected to stage movement. I’m having some issues with the image, I’m unable to get an in-focus preview at all, even after auto-gain setup. The illumination setup is weaker for some reason, compared to my x60 objective, even though that one uses the same photography diffuser, and the LED is glowing strong when I take it apart.
I can move it closer to the sample, but then I lose it completely and just have a completely white image. I’m using immersion oil in between the objective and the sample. I put some on top too after a while, but my condenser isn’t touching it, so I don’t know if it does anything at all?
Anyway, here’s the default speed moving the OFM around, on x100 magnification, with bad lighting and blur. Shows movement though
Quite a few updates, some of which I’ve shared in the thread about nema14s. The new thing about it - I have added a touchscreen. Had to hack away at the plastic of the tray near the hdmi connector, the print doesn’t allow for the microHDMI to make proper contact. I’ve created an issue against this in the Microscope gitlab.
The screen looks great, though it was annoying to setup. The controls are small, but usable. I have encountered a bug with the setup though, without a keyboard it seems I’m unable to skip the calibration phase. In Mobaxterm via VNC I’m able to press “escape” to avoid it, but no way to do that with just a touchscreen. Maybe pressing outside the modal window should close it, or a button that says “remind me next startup”?
The stepper motors are now integrated with the OFM server software so manipulating the microscope is possible. Below is a vid from the Contributions thread about the steppers on how this looks.
