749
2021. JAN. 18.
A proper introduction and some details of what to expect on this blog. The beginning is the most important part of the work. ― Plato, The Republic.
Discalimer: The blog has nothing to do with meerkats. I have the header image as my wallpaper everywhere for Years now and had no better idea. (had tbh, but went full retard and did not purchase ledtools.com immediately after lookup and was ofc sale listed in a couple of hours from godaddy..)
This is the very first post on the blog and so I think a proper introduction must be in place of what to expect and why this all was started. As for the latter, as highlighted in the site footer, mainly in order to share some of the tools and projects that I had on shelf for quite some time now, hoping that it might be useful to someone. Since most of these have been terminated when reaching a point where I either lost interest or was already satisfied, I still have high hopes that someone will find inspiration, maybe solution to a problem or even catch up where I left off and make something useful out of these contents. I have some plans moving forward and the fate of Meerkatblog greatly depends on readers interest. Currently It's really just about getting rid of my demons of unclosed and never released ideas.
So lets see what exactly are these projects.
This blog itself is one such idea of mine. It started with the LED color mixing tool that I originally designed to be a single content website. That was long ago with php + frontend JS and I was never really satisfied with it. Then came 2020s EoY vacation with 2 weeks of isolation in the COVID situation that resulted in this. Since I wanted to release early 2021 rather than further polishing - no secret here - is under construction. I'm aiming for adding 2 more Meerkat Tools this Year and post as much as I can. Commenting is still WIP, just as well as authentication. The latter will be handy with the coming Meerkat Tools.
Since some of the content here was a struggle to realize, I'll make tutorials + GIT commits about the more interesting and unique parts.
The tool originates from an Excel sheet that I'll also share in the future. Nothing to do with it from my side, besides building a frame around on the blog. If You are amongst the lucky ones, it could have begotten on symposiums or by me - or whoever shared thereafter. Since I'm out of the lighting business for a while, I'm getting regular requests of updating and maintaining the database. Well, that's practically impossible now time wise. Hence the idea of a web based tool, where the community will be able to add / edit LED models. I'm super curious how this will turn out. Current LED models are from datasheets, CREE PTC tool, OSRAM network and Lumileds also had a closed customer portal but I dont have access to that any more.
Some major updates are also coming here. Most importantly a horticulture calculator, a HCL calculator (CS calculator from Lighting Research Center) and the community interface for registered users ofc. On interest, I'll advance then towards color deficiency simulations, but that's still far far away. BTW, does anyone have the TLCI samples?! I also want to add that.
Working as an optical engineer has its challenges for sure. To claify, I was the one for long who designed the optical surfaces around the light sources that then concluded in a system capable of satisfying standards requirements. This is more or less similar to a 3D artists workflow. We create 3D surfaces (parametric), some more math involved with less art content - and comes the simulation which is practically the same raytracing that 3D rendering does, framed in a Monte Carlo simulation with accurate optical models. The problem here, especially with outdoor lighting is that the goal is not always clearly defined. Meeting the requirements is one thing, but acheving something that is good is still a long way ahead and the road is not paved. If You change one paramter, every illumination and luminance result will change in a non-linear way. Mayhem.
Enter our protagonist here, the LIDViewer. Originally I started it to have a tool to quickly review the simulation results with. LID = Luminous Intensity Distribution, Viewer = Viewer. The goal was a fast workflow. (FAST. unlike Photometric Toolbox) Then came a wierd idea: once I can tell from an intensity distribution how well it will perform (standard, glare metrics, optical efficiency..), why not generate the ideal output for a lighting scenario with machine learning and AI? It works. With this, optimization of the optical surface gets a whole lot easier. We dont need to tune parameters on a chaotic design table, only minimize the residual compared to the 'ideal' light output. Those who design optics sure understand the meaning of this.
Then time passed and features grew. I've shared to software with colleges here and there but have never formally released. Meerkatblog will shelter a home for the LIDViewer, its source and its companions. Also, WPF is an infinite source of technical tutorials. There will be categories about WPF MVVM direct 2D and D3D11, custom charts, embedded views and a series about managing files for sure.
Not a typo. If You know STM32 CubeMX and also know the hassle of digging out ports properties of the Espressif microcontrollers, then You have probably already figured out the point of this tool. Will create a Meerkat Tool here and You will also be able to install as a PWA. Currently implementing ESP8266, ESP32 and ESP32s with the most common packages (WROOM WROOM).
I love 3D printing. FDM especially. There will be 3D printer builds (all custom), benchmarks, component reviews and some challenging prints. Looking forward?
This is an ESP32 project. When the lockdowns started I bought a load of racing sims on humblebundle, only to realize that I suck. Big times. I obviously must have had found something to blame in this situation, which happened to be the controller. (practice makes perfect probably, though really unconfortable for me). The idea was that racing wheels are, well not hassle free, to say the least. But what if the same feeling of precise steering could be achieved with a controller imitating a wheel + possiblilty of attaching pedals. I made a quick MVP with an ESP devboard, MPU6050 breakout and some pots. And it worked! Windows sees ESP32 over bluetooth - can even set parameters like battery level and device state (not implemented by some major controller brands - wgrrr). So currently I have a layout on a 4 layer PCB, consisting an X-Box like layout and the gyro sensor. It has SSD1304, buttons and joys. There is a holtek USB HID IC, since ESP32 (non s) doesn't know that, a I2C battery charger (see below #SmaCha) and tuned opamps for x-y trackpad (yes!). #Gacco will be a complete infrastructure of LEGO components so everyone can build a custom Frankenstein's monster of a controller and then the best part; there will be an online configurator that can create STL file for 3D printable housing for it.
Is a smart battery charger breakout board with the BQ24298 IC. Has I2C for setup and control. Has long push power-on button feature and can be turned off with uC GPIO. I honestly dont know why this is not mainstream. BAT temp thermistor monitoring (JEITA profile), BAT temp thermistor monitoring (hot/cold profile), IC thermal regulation, ICO (Input Current Optimization), IINDPM (Input current limit), Input OVP, Integrated ADC, Power Path, USB OTG integrated, VINDPM (Input voltage threshold to maximize adaptor power)...
ESP32 + TCD1304 + 600 line grating in a 3D printed housing equals big win with tons of fun. There is an inherited problem with almost all the handheld spectrometers that there is a Hamamatsu Micro-spectrometer IC inside that has a low resolution. There are techniques to "enhance" the signal with priori assumptions about the sources but still really inaccurate. I've RFQed once the IC and remember something like 400 USD each. (why the hell does Hamamatsu market this as something one will build into an office lamp??) Using the usual linear CCD approach requires a great deal of quality control to even work but still, serves a better quality at a lower price. I think the single IC solution got widespread due to its ease of use and to my surprise, there is a great deal of mountebank business going on in this market. #ESPectrum has currently an Android app (Oreo flavor if I remember correctly) that I'd like to port as a PWA. Also a great plaftorm for many future projects involving spectrophotometry and still photography.
This one didn't got quite as far as I wanted to after I realized my serious lack of knowledge in music theory but I definitely want to have it cleaned up once. ESP32 again. one I2S is input stream, one is output. 24bit, HIFI quality. Guitar goes in, distortion goes out. Bypass works for now and on the long run, I want an open source library and a community here to configure presets. Control will be through WIFI or Bluetooth. Additionally some Rocksmith like app would be awesome. Why doesn't it already exist?
This is yet another ESP32 project for horticulture lighting. Basically an advanced light sensor with AMS color sensor ICs. I made two variants; one sees tristimulus and the other one has 18 discrete spectral channels. Communication is over WIFI. Can log and control nodes for spectral controlled horticulture and floriculture. The layout had some major issues so I'll fix it and also add some additional sensors.
Another Meerkat Tool here. Grabchart lets You pasting an image - preferably a chart on a canvas. Then poins can be defined and interpolation / regression can be set up with MANY variables. This is still under development, though would have been a huge help with some LED characterizations for the LED Mixing Tool. Yes, I know that there are quite a few robust solutions out there so the killing features in a nutshell:
There's a Patreon link in the footer section that I encourage You to use. I have no idea at this point if anything here could be sucessful from an audience demand point of view so the site is starting off with the minimum possible expenses available for hosting and operation. In case the patreoning will have any success, that will cover the blogs expenses and if in an extreme case there might be any profit; that will turn into charity within a presentable format. 2021 will be the Year of orphans here.
And finally. I will give a try vlogging these contents on Youtube. Link in the footer. Nothing available yet.
Now, without any detailed explanation, I've also set up the following tags in advance:
I'm thrilled that people might read this :)
Have a nice day.