Hello. The optical principles are as described in the videos - you need to project an image of your computer screen to be focussed on the same plane as the ocular lenses' focal plane. Achieving this with other microscopes can be a problem because their 'trinocular' ports point down to the specimen so if you shine an image down there then part of your projected computer screen image will also go down onto the specimen and may cause unwanted back-reflections. The PUMA arrangement allows the image of the TFT screen that gets past the beam-splitter to go horizontally and be either absorbed by a light sink or used in the PUMA trinoculular port which is lateral (at 90 degrees) to the specimen. So this image does not spoil the specimen. I suppose you might be able to achieve this on a standard microscope by using a double header-type splitter port or a port that would otherwise be used for a pointer but the details of getting the projection and zoom right might be tricky - I have not tried.

​@@PUMAMicroscope Thank you for your comments. I thought that my trinocular head had a built in beam-splitter but after reading your comment and checking this might not be so. I will have to check in detail. I suppose the alternative is to adapt an attachment in between the binocular/monocular eyepiece similar to the puma so I can include a beam-splitter into it (you use a SSM instead of a prism here it seems). My very loose understanding of optics (from school many years ago) suggests that I will need to add a focusing lens to this extension attachment to account for the extra distance made (as the microscope is a 160mm one as it is not infinity corrected). Would this be a correct general outline of the task at hand? Thus in this case I would (if you allow me to) adapt your Puma system to include an additional focusing lens and also adapt connectors at both ends to join this to my microscope. Would this be OK with you. I would also appreciate any additional pointers you could give. Thank you!

@@letshin Your trinocular head does indeed have a beam splitter in it, just not in an optimal configuration for use as a AR projector set up. Whatever method you try, you will need to use lenses to ensure a projected image of your TFT screen (or smartphone screen if that is what you are going to use instead) is focussed onto the focal plane of the oculars and avoid getting light from the screen to go down onto your specimen (ideally).

@@PUMAMicroscope Thanks, I think I would like to start off by making the PUMA AR projector and the above stage optics to figure out what I need to do to adapt this for my own use. I checked the Github and looking through the 3d printing guide it seems I will need the binocular head (BN_) files as well as the AR projector (AR_) files. Will I need the filterblock module as well? Is there a document/decision tree that relates how the different modules link together? Edit: The possible configurations are at the end of the BoM Excel file so I'll refer to this instead. I've ordered the required mirror and beam splitters from these (I have spare eyepieces and objectives) so will see how this goes. Thanks!

@@letshin You do NOT need the binocular head for the AR projector. You DO need the advanced filter block. Some people have told me they have had difficulties getting the concave lens for the AR projector. It would be interesting to know if you can get it.

Hello, Re 'many parts' - I do agree, but there is a reason for that. Also, this, plus the time and work involved in putting it all together is why is costs so much to buy (instead of make DIY) - so can't compete with off-the-shelf prices for mass-produced standard scopes like of Swift. Re 'flimsy' - I respectfully disagree. Don't get me wrong. PUMA is not a stable heavy metal desk scope - I do admit that in the videos - but then it was not intended to be such and, furthermore, you can stabilise it to a desk and use the stabiliser module to increase stability as I demonstrate in my video on the stabiliser here: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-d8tMtWFhNY4.html ). However, PUMA is not primarily designed for people who 'just want a cheap small scope' - if that is your goal then sure, I agree it would be better for you to just buy a standard scope. PUMA is for those who need many options and complete customisability (not just modding the condenser of a Swift to give you dark field) and portability (so that rules out Echo revolve). For example, the Köhler Fourierfeldblende projection, augmented reality, Z-motor, epi-illumination in bright or fluorescence, epi-polarisation and trans polarisation, coherent light microscopy, etc. - these modular options, and more, can all be bundled into a little light bag with PUMA and taken along on your journey. This is something you can't do with a Swift or similar - but then if you don't need any of those things with ultraportability, then you don't need PUMA. Also, PUMA is for teaching and learning about how microscopes work and for experimentation with new designs and optical modalities - like interference microscopy, structured light illumination, superresolution microscopy, etc. (which I have not done videos on yet - but keep watching). Regarding a heavier base, I did consider making an extra stable base option but not to be filled with concrete (most people would not want to work with that,. esp. in small quantities), but rather a 3D printed stand with a base that may be filled with wet sand or clay or pebbles (whatever is out in the field where you are working) so you can have a heavy base on site without having to carry the heavy weight along with you on your travels - just empty the base container when you finish and move on to the next site with a nice light scope in your bag. I am also working on a more stable base for XYZ precision motor-controlled microscopy - more of that to come. Remember PUMA is not just a standard microscope made with 3D printing and DIY - think about what each of the letters mean and put them all together. How many of the other scopes you mention have all those things (P. U. M. A.)? But not everybody needs or wants those things - so, for sure, PUMA is not for everyone. Thank you for your comments. PJT

Thanks for the feedback. There is additional info in this sequel you might also like: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Ai86SMBJqr8.html

Thanks. You may also like this video where I explain in a little more detail the 'conjugate planes' aspects of Köhler and how it relates to what you have just seen: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Ai86SMBJqr8.html

Thanks.

Thanks for the comment,. Clearly I can't know for sure but here are some possibilities / suggestions: * One or more of the programs may not have installed correctly. Try a re-start of your computer. If that does not fix it, try uninstall and re-install +/- restart. Try typing the commands 'deconvolve' and 'fitsdou' (without any arguments) on a command window and see if they are recognised (you will get the standard usage message if it is recognised). * It may be that some antivirus software has decided to block or quarantine some of the programs so you might need to look into that and re-install once permissions have been granted / files un-quarantined. * All BIPS programs cannot handle any file name or directory path that contains spaces - so if you installed the program to something like 'My Programs' or if any of your files are in folders like 'My Document' (or anything with a space anywhere along the full path) then this could cause some programs to fail to find files and run. * Check any feedback messages on the console window - these might give you a clue as to where the problem occurred or even what it is. If you are using the standard demo without alteration you should not require any additional dlls (if you tick the 'Use parallel' box then certain multithreading dlls will be required). If you manage to get it working, consider making a reply to this comment to state what the problem was and how you fixed it to help others who have a similar issue. Thanks,

Thanks for the feedback.

Thanks. I will be doing more deconvolution later. Next up will be my series on the Fourier transform - so stay tuned for that.

Thanks.

Hello. Sure - what you have said should be possible. There are a number of ways of achieving it. Let's assume you have built the AR Projector optical tube, then: One method is to use a smart phone or tablet in place of the standard TFT screen - for example holding the phone at the right position in front of the stand in the same way I held up the slide of an eyeball in my video on 'Photology 5' at about time point 00:10:12 ( ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-u_0xczp4210.html ) with some drapes to shield from outside light. Then simply use this phone or tablet as a wireless screen 'mirror' or 'screencast' device to receive the video signal from the phone or tablet your students are holding. The other option is to use the standard TFT screen AR projector but you will need to write the interface yourself because that is not a standard feature of the current PUMA Control software - and you may also need to use a more powerful Arduino than the standard 'nano' version I use in my demos here (perhaps like the Arduino Mega). You would then connect the tablet or phone to the Arduino with your own software painting to the TFT screen in response to your students touch-movements on the tablet / phone. I haven't done any of this so can't guarantee it will work but those are the ways I would approach it.

Thanks. I'm currently working on the Fourier videos.

Thanks. Yes you can use PWG to measure BFP - although simply pointing the objective at a distant scene will also work. I will be using the PWG to demonstrate Abbe's diffraction theory of image formation in the light microscope and some Fourier optics work (at least that's the plan). As for white light interferometry and OCT - I would like to make my precision PUMA XYZ stage a reality first but, if successful, those are some of the applications that the PWG can be used for in conjunction with the beam splitter of the advanced filter block (although it might also work with the Z-motorised current stage with precise microstepping in Z - I need to experiment). So there's a tonne of PUMA material still to come.

@@PUMAMicroscope This looks very promising. I'm very hyped! I hope you'll find some good solutions for the motorized stage, but it's really not a trivial solution. Atm my best result was using an aluminum frame made of 2020 extrusions to mount the microscope on a linear rail using a lead screw for the Z axis, and two double linear rails (motorized) for X and Y. It's surprisingly very precise and works even with a 100x objective. But, well, it's probably larger than what you ever want (something about 24x24x40cm - 40 cm height mainly because I have a 80mm tube extension). But on the other hand, it's very lightweight (the frame contains more emptiness than anything...) so I have a lot of vibration issues. I'm not sure if motorized and "portable" are really in the realm of the possible. But linear rails are pretty good at this task, and even if the precision is not perfect, if you are doing backlash compensation you can get great results and easier stitching later. Next I want to try to 3d print linear rails to see if we can get good results... or not (in order to cut prices)!

@@PUMAMicroscope Also, it may interest you - I recently got a monochrome global shutter raspberry pi camera based on the OV9281 sensor. It's only 35€, and the quality is really good - I think it may be a great deal to use for microscopy in specific applications, because monochrome camera are generally VERY expensive, especially with a global shutter... If you want to get one just search for Innomaker OV9281 camera.

@@PUMAMicroscope Also, it may interest you - I recently got a monochrome global shutter pi camera based on the OV9281 sensor. I think it may be great to use for microscopy in specific applications, because monochrome camera are generally VERY expensive, especially with a global shutter... Search for Innomaker OV9281 camera. You also have USB ones with very low framerate on ali. But probably good enough for experimentation.

Also, it may interest you - I recently got a high framerate monochrome global shutter pi camera based on the OV9281 sensor for very cheap. Only "downside", it's only 1280 x 800. But I'm sure it's enough for a lot of applications. I plan to use one to make a spectrometer, and another one for microscopy in general.

Thanks. Yes - micro-Z stepping interferometry of chip surfaces is one application. I would like to try that and make a video on the results - but of course I'll be very happy if others beat me to it!

Thanks for the encouraging comments. Sometimes I do wonder - I mean 'big' RU-vidrs can just post a video of them sneezing in front of the camera and get 10K views in the first hour - it took my first PUMA video almost 2 years to hit 10K and most of my other videos never reach 1K. But I still have some knowledge to teach - which I think at least some people will find valuable - so I will carry on for now.

​@PUMAMicroscope Big RU-vidrs videos have high entertaining value. Yours have tremendous value but for niche audience. And i'm grateful they do exist.

​@@PUMAMicroscope We might be a small crowd, but we're a crowd of enthusiasts, professionals and scholars all over the world that are highly interested in the topic.

@@PUMAMicroscope It's a bottle in the sea waiting to be found. On youtube the quality is not what get more likes, it's what the algorithm promotes. I experienced this myself with a simulation I made, requiring a lot of work. Then I published the video and I got a few likes... Months later, for whatever reason, the algorithm decided to highlight the video. It got 240k views and thousands of likes in just a few days... But in the end, it didn't bring me anything at all, because youtube is youtube. It's mostly for entertainment. Your channel may be small, but people here are very interested. Myself I watched more than 10x some videos just to understand some concepts. Optics is really hard to grasp without visual / real world applications. And you did a great job with that. You really are the best / most comprehensible reference on the subject, IMO.

@@oni2ink Thanks.

Yes it is the LC_Cap (p.142 of current 3DP guide PDF on GitHub). In the video clip the Pol_Adjustment_ring has an LC_Adjust_collar fitted into it (p.111 of the PDF) held by the three thumbscrews and this LC_Adjust_collar provides the female thread that the LC_Cap threads into.

@@PUMAMicroscope Thank you very much, that makes a lot of sense. Great work btw! One more question both monocular tubes (CM and for regular ocular) have very thin thread wall that screws into filter block, but there is a lip inside that makes this hole narrower anyway. Is it done on purpose, or could it be thickend? (I just broke off the thread off one of the tubes :()

@@lextorn92 This is a potential weak spot - at least one other builder has had the same misfortune as you. The lip is a deliberate design feature - it is part of the anti-reflective baffle system and it needs to jut out like that to work - so just making the wall thicker is not a solution. It might be that I need to adjust the infill density (or make it 100%) for this part of the tubes. In the mean time, the most practical solution is just not to tighten the tube too much - you will have gotten a feel for how much tension it can take from your unfortunate experience.

Hi. Sorry your latest comments were filtered out by the YT auto-filter due to containing a link (my channel is set to disallow links in comments other than my own) but your suggestion of buttressing the inner diaphragm may work - I would need to try it. Thanks for the suggestion.

@@PUMAMicroscope No worries, thank you!

Yes and Yes. So in the first case I have done standard Zernike phase contrast with the 1W bulb using a Leitz phase objective and the PUMA SLM as condenser phase ring and it works very well with good brightness - and that is with the SLM which absorbs a lot of the light so it would have been even brighter if I had used a plane phase mask in the IAD slot instead (e.g. 3D printed or hand drawn on a glass coverslip). As for the second question - you can get a 3W version of the same footprint LED but you will need to make or get a higher wattage power supply than the standard PUMA power supply for it and also a much better heat sink than in the design shown in this video. Here is a link to a shop that sells them: www.aliexpress.com/item/32881415165.html . One of my goals for future development is to improve the lamp housing design, heat sink design and power supply in general - especially the heat sink because it is very crude, even for the 1W bulb - I certainly would not just plonk a 3W bulb into this design case, the heat would probably melt stuff.

Thanks.

Thanks. I'll have more to say about Schlieren at some point in future videos.

Your guess is correct - the oval, projected at 45 degrees, gives a circular area of illumination. However it is not about 'as seen in the eyepiece' but as illuminating the back focal plane (BFP) of the objective so as to get maximum resolution and minimise unwanted phase contrast. You could use a round mirror if the minimal dimension of the projected oval (because the projection will be oval with a round mirror) also fills the BFP of your objective. This should be fine for low power objectives but the higher you go the more likely you are to get issued with poor quality illumination if the mirror is not large enough. So, with a round mirror, to get equivalent quality of illumination for high magnifications, you will need not only to alter the shape of the receptacle but also its size and a bigger size may not rotate to 45 degrees unless you also elongate the legs. Having said all that, you might actually be able to 'get away' with using an oval piece of aluminium foil instead of the mirror. As long as you are not using this as part of a Köhler projection system, the slight diffusing effect of the aluminium foil might actually improve the quality of illumination compared to a clean / pure mirror.

Thanks for the insight. I am self taught in using CAD and come from a medical (not engineering) background. This might explain my unusual approach. I used 'Part Design' for some of the PUMA components though.

@@PUMAMicroscope The part workbench is less common but definitely not impractical since it's the original modelling method FreeCAD utilized. Your work on this device is impressive and I've loved the scope I bought from you. I'm getting ready to upgrade it, which is why I'm going through your videos at the moment.

It's called a 'slide mailer' (a single slide mailer in this case because this type holds just one slide ). This is what we use to send glass slides in the mail - hence the name.

@@PUMAMicroscope I know the generic term 'slide mailer' but when I search I only find the hinged opening type. This is like a sliding style of mechanism to it. I did search for this specific type of opening and could not find it. I am able to find a generic slide mailer that hold one slide but I am in search for this gliding style slide mailer. Please help if you have a link for this specific style of slide mailer

@@user-cy2sz5kr4l Oh - I see. I get these from a company called CellPath Ltd. They call it the 'SLIDERITE I MAILER' and product code is 'BCR-0100-02A'. Hope that helps.

Thanks for taking the time to watch and comment.

Thank you. I did a lot of work on super-resolution in bright field microscopy years ago and I hope to bring some of that to this channel over time. Yes - I hope also to demonstrate some SIM with PUMA - but there are a several more videos on the theory side I want to get out first so everyone can join the party.

@@PUMAMicroscope That sounds phenomenal! Thank you so much for continue to share your knowledge.

Thanks.

Hello. I can't give a definitive answer since I don't have direct experience with these variants. If it were me I would proceed by first measuring the front and back focal lengths of the lens and compare that to the specs I give in the Köhler PDF on the GitHub page. If the focal lengths are not too far off you should be able to use them by adjusting the 3D print models. Some empirical experimentation may be required for fine adjustments (tube lengths and separations, etc.). Be guided by the optical principles and results. Sorry I can't be more definitive but without such lenses to play with I have no more specific / direct experience to offer you. I'd be interested to know you results.

@@PUMAMicroscope Ok thank you! I'll do some measurements / tests when I have some time and I'll report here. Do you buy your lenses from the same place or use the same old stock?

@@oni2ink I bought a load of them some time back (from the link in the BoM) and still have some left so haven't bought any recently,.

Thanks. Stay tuned for episode 5 in a couple of weeks. Very few people are watching these videos - maybe they are too technical so only for specialist audiences? Either way, I'm glad some people appreciate them.

My understanding is that Schlieren can be seen as one particular type or variety of 'oblique illumination' but not all forms of oblique illumination are Schlieren. For example simply shining a light obliquely on a specimen (as I show in my video on epi-illumination here: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-cAEB10K8PqI.html ) is a form of oblique illumination microscopy but it is not a Schlieren method. In the olden days, many microscopes had condenser / mirror holders that could be swung from side to side to give oblique illumination from below - again providing oblique illumination in a general sense without the Fourier-pane bisection characteristic of a Schlieren method. If you would like an overview of these things do a Google Scholar search for a paper entitled: "Beyond Brightfield: “Forgotten” Microscopic Modalities" by Radek Pelc. I hope that helps.

@@PUMAMicroscope Thank you, I'll look into it.

Yes - I tried to show that with the dotted lines which mark out one specific arc - but I agree I could have made this clearer. For anyone reading this: The line length 'g' is the phase difference between yellow and white wavefront lines and is this length g is the same at one end of the arc marked by the dotted lines. At the other end of this arc the phase difference is always zero between those yellow and white wavefronts (because here the yellow and white lines cross) so the change in phase difference (g minus 0) is the same despite the arc between the dotted lines getting longer and longer as you go further from the wave. I hope that makes it clear.

Hello. For any infinity objective (regardless of its make or other features) when the specimen is at the correct working distance from it, the imaging rays coming out the back will all be parallel rays (the imaging BEAM will be diverging - but never mind about that for now, it does not affect the answer to your question). To get those parallel rays to converge to a focus so you can see the image in focus you need to get them to converge at the focal plane of your eyepiece. In the PUMA microscope there is only 100 mm from the position of the (optional) tube lens to the focal plane of the eyepiece, therefore the tube lens must cause those parallel rays to come to a focus 100 mm away from it and so, by definition, it must have a focal length of 100 mm. Most professional infinity microscopes have tube lenses of f=200 mm or thereabouts. This is why using the PUMA with its 'standard' infinity tube lens of f=100 is a compromise and does not give good image correction (the objectives are designed for a tube lens with much longer f). However, you could use a tube lens with f=200 (or just over 200, something like 220 mm) and then you would need to increase the distance to the eyepiece by the appropriate amount - so if you use a tube lens with f=200 mm , just make the ocular tube longer by 100 mm (although things could get a little 'wobbly' if you stick a camera on the end of such a long tube without appropriate support). I hope that all makes sense.

@@PUMAMicroscope Oh I see! Thank you very much for the explanation. Do you think it may be worth the effort to use mirrors to make the system more compact? Or it may just reduce image quality for little gain?

@@oni2ink That mirror thing was the question I had to ask myself when designing the canonical PUMA. I decided against it because it would mean using first surface mirror or a TIR prism and would increase the cost and also mean singificantly more 3D printing and construction. However, for any individual, like yourself, if you are building a system for your own use then the decision is up to you. It shouldn't reduce image quality significantly if you use one first surface mirror but it will add weight and the system will have increased bulk which, if not designed carefully, cause increased vibrations thereby cancelling out the desired benefit. It may be easier to just design a more stable straight monocular tube (thicker walls with buttressing, etc.). Over to you to experiment and see what works.

@@PUMAMicroscope I see, thanks for the clarification. I'll try to increase the tube length, and the thicker walls are a neat solution. My Olympus BH2 have a magnification changer just above the tube lens. If I want something like this on the infinity PUMA version, do you think I can add an optional achromat lens on the tube? For example by using something like your filter slot system?

@@oni2ink For that you will could use a beam expander between the objective and the tube lens. This might reduce image quality - I would avoid it and use a higher res camera instead.

It is 'almost' possible by using the DI_M3_Adjustment_ring model (page 124 of the current 3D printing guide PDF) to hold the illuminator. This is normally used to hold the lamp for the epi-illumination config, however if you can print a short male thread ring to act as an adaptor you can attach the DI_M3_Adjustment_ring to the base of the condenser (where the DI_Cnd_to_UC or DI_Cnd_Mirror_holder_socket would normally go). I say 'almost' possible because I have not included such a 'short male thread ring' in the project as yet but you could fashion one using the top male threads from those latter-mentioned parts' FreeCAD models. I did actually make such an adaptor but just haven't released it in the GitHub repo because I did not see the advantage or need to stick the illuminator directly under the condenser - it will not give good illumination like that (poor NA). In a future video I will show you the PUMA High NA illuminator which has a similar compact config to what you want but uses a specially made diffuser in place of the Abbe condenser to give high NA illumination.

Still dope but yea i was shopping for a scope to today and was floored by whats avaliable for under 400. Any suggestions on bramds and where to buy?