The smearing on the image could be due to the fact that you don't use a shutter, so you see brighter stripes under bright areas of the image as you still iluminate these pixels while the sensor data ist shifted out towards the top. I experienced this effect back at university with a LN-Cooled CCD for Spectroscopy. The stripes disapeared as soon as you used the shutter instead of disabling it in the open position (but fokussing at 100ms integration time and continuous readout with a focal plane shutter isn't much fun).
I had an "Oh SHIT!" moment when I realized that the weird trapezoidal shape of that light guide was for keystone correction of the light source. Very clever.
The pricing on that sensor is about right, I looked in to these many years ago when they were still in production since it's the only large sensor you could actually buy. Really cool to see one in the wild.
The green LED looks different from the others and uses phosphor conversion because of the "green gap" problem where green InGaN emitters suffer efficiency droop at high currents. Phosphide based emitters don't start becoming efficient until around 600nm so also can't be used for high power green emitters. See the paper and plot by Matthias Auf der Maur in his 2015 paper on alloy fluctuations in InGaN as the cause of reduced external quantum efficiency at longer (green) wavelengths.
this is exactly the kind of nerdy deep dive one would expect on this channel - thank you for sharing. Not just a throwaway comment or opinion...but a fact, with a paper to back it up. awesome!
Narrowband filters we use in Astronomy (Astrophotography) are sided- they work best passing light in one direction so I guess the arrows on the filter frames indicate which way round to install them in the filter wheel.
The red LED reminds me of one from Roithner Lasertechnik. I have a Symbol 2D scanner which uses two very bright LEDs from that company, one red and one red-orange. The red-orange is behind a lens which focuses it into an extremely narrow beam.
Thanks Mike. Ooof! - with the level of bodgery going on around 15:48 I think shame would have made me do a board re spin, out of my own pocket if I had to.
Good god, I wasn't prepared. The little smd going from the middle leg of a 3-pin chip is hilarious. Then we get to the silica gel packs and the antistatic bag.
Maybe the camera should not be illuminated during readout. From the datasheet of the sensor (Onsemi): saturation 40300 electrons, read noise 12.5 electrons per pixel @ 18MHz (quite bad). quantum efficiency 62% (if it has micro lenses), frame rate 1 Hz. lateral overflow drain to prevent blooming protects against 800x (factor increases linearly with exposure time) saturation exposure (32e6 electrons per pixel at 4ms exposure time), microlens has +/- 20 degree acceptance angle i guess it would be good for astrophotography
*Summary* * *0:00** High-End Optics:* The system utilizes heavy, high-quality lenses and mirrors for precise imaging, weighing around 4 kilos each. * *0:49** Narrow Band Filters:* A filter wheel with five narrow band filters (488, 525, 570, 630, and 700 nm) ensures accurate fluorescence detection and rejection of excitation light. * *2:01** Customizable Illumination:* The Lumen Dynamics light source offers five individually controllable LED wavelengths (430, 475, 520, 575, 630 nm) with varying power outputs. * *3:45** TTL Control:* The light source is controlled via a simple TTL interface, enabling easy on/off switching for each LED color. * *12:55** Sophisticated Camera:* The system includes a 50-megapixel Kodak KAI-50100 CCD camera with a Peltier cooling system for reduced noise. * *14:54** High-Speed Data Transfer:* The camera features dual analog-to-digital converters to manage the high data throughput of the 50-megapixel sensor. * *18:11** Possible Issues:* The video creator noted some potential issues with the camera, including image smearing and a limited dynamic range. * *11:45** Low Runtime:* Internal data suggests the system has seen minimal usage, with only 20 minutes of recorded runtime for the green LED. * *20:38** Availability on eBay:* Both the illuminator and camera are expected to be listed for sale on eBay. *In short, this video details a teardown and analysis of a high-end fluorescence imaging system from Fluidigm Polaris, focusing on its illumination source and camera. While potentially powerful, the creator notes some possible issues with the camera and indicates limited prior usage.* i used gemini 1.5 pro to summarize the transcript
The whole sensor module incl. a 5 grand 50mpix sensor for 49 £.. highest bid atm - Though also a limited CCD sensor, but for the right buyer its a steal at these relative low sums. - Architecture Full Frame CCD (Square Pixels) Total Number of Pixels 8304 (H) × 6220 (V) = 51.6 Mp Number of Effective Pixels 8208 (H) × 6164 (V) = 50.5 Mp Number of Active Pixels 8176 (H) × 6132 (V) = 50.1 Mp Pixel Size 6.0 m (H) × 6.0 m (V) Active Image Size 49.1 mm (H) × 36.8 mm (V) 61.3 mm (Diagonal), 645 1.1x Optical Format Aspect Ratio 4:3 Horizontal Outputs 4 Saturation Signal 40.3 ke− Output Sensitivity 31 V/e− Quantum Efficiency KAF−50100−CAA KAF−50100−AAA KAF−50100−ABA (with Lens) 22%, 22%, 16% (Peak R, G, B) 25% 62% Read Noise (f = 18 MHz) 12.5 e− Dark Signal (T = 60°C) 42 pA/cm2 Dark Current Doubling Temperature 5.7°C Dynamic Range (f = 18 MHz) 70.2 dB Estimated Linear Dynamic Range (f = 18 MHz) 69.3 dB Charge Transfer Efficiency Horizontal Vertical 0.999995 0.999999 Blooming Protection (4 ms Exposure Time) 800X Saturation Exposure Maximum Date Rate 18 MHz Package Ceramic PGA Cover Glass MAR Coated, 2 Sides or Clear Glass -- Features ⤵ • TRUESENSE Transparent Gate Electrode for High Sensitivity • Ultra-High Resolution • Board Dynamic Range • Low Noise Architecture • Large Active Imaging Area Applications • Digitization • Mapping/Aerial • Photography • Scientific - Thx for the tear down Mike, always a joy
He lives! ❤
Ever heard of being able to record video and play it back later?
The smearing on the image could be due to the fact that you don't use a shutter, so you see brighter stripes under bright areas of the image as you still iluminate these pixels while the sensor data ist shifted out towards the top. I experienced this effect back at university with a LN-Cooled CCD for Spectroscopy. The stripes disapeared as soon as you used the shutter instead of disabling it in the open position (but fokussing at 100ms integration time and continuous readout with a focal plane shutter isn't much fun).
I didn't think of that, but makes sense
Thanks Mike. Your videos are always interesting.
given the cost of the CCD you think they could have run another PCB for it
Definitely the best 10-second pinhole pictures I've seen
15:49 that is some great bodging on of caps, they really didn't want to respin that PCB huh
Great vid - gets right into subject in title, its packed with information, wraps up quickly. Should get a YT award! imho
I had an "Oh SHIT!" moment when I realized that the weird trapezoidal shape of that light guide was for keystone correction of the light source.
Very clever.
Was depressed today and then a new mike video dropped and now I’m genuinely happy to get my tear down fix
The pricing on that sensor is about right, I looked in to these many years ago when they were still in production since it's the only large sensor you could actually buy. Really cool to see one in the wild.
The green LED looks different from the others and uses phosphor conversion because of the "green gap" problem where green InGaN emitters suffer efficiency droop at high currents. Phosphide based emitters don't start becoming efficient until around 600nm so also can't be used for high power green emitters. See the paper and plot by Matthias Auf der Maur in his 2015 paper on alloy fluctuations in InGaN as the cause of reduced external quantum efficiency at longer (green) wavelengths.
this is exactly the kind of nerdy deep dive one would expect on this channel - thank you for sharing. Not just a throwaway comment or opinion...but a fact, with a paper to back it up. awesome!
Heavy optics indeed... Spare no expense, cost no object. Splendid build quality. The CCD is a thing of beauty!
Babe wake up a new mikeselectricstuff has dropped!
Narrowband filters we use in Astronomy (Astrophotography) are sided- they work best passing light in one direction so I guess the arrows on the filter frames indicate which way round to install them in the filter wheel.
The red LED reminds me of one from Roithner Lasertechnik. I have a Symbol 2D scanner which uses two very bright LEDs from that company, one red and one red-orange. The red-orange is behind a lens which focuses it into an extremely narrow beam.
are those two ram chips stacked on top of each other? 13:47
yaaaaay ! more overcomplicated optical stuff !
Why overcomplicated?
Thanks Mike. Ooof! - with the level of bodgery going on around 15:48 I think shame would have made me do a board re spin, out of my own pocket if I had to.
Good god, I wasn't prepared. The little smd going from the middle leg of a 3-pin chip is hilarious. Then we get to the silica gel packs and the antistatic bag.
Maybe the camera should not be illuminated during readout.
From the datasheet of the sensor (Onsemi): saturation 40300 electrons, read noise 12.5 electrons per pixel @ 18MHz (quite bad). quantum efficiency 62% (if it has micro lenses), frame rate 1 Hz. lateral overflow drain to prevent blooming protects against 800x (factor increases linearly with exposure time) saturation exposure (32e6 electrons per pixel at 4ms exposure time), microlens has +/- 20 degree acceptance angle
i guess it would be good for astrophotography
$20 thousand dollars per minute of run time!
Did you transfrom into Mr Blobby for a moment there?
ironic that some of those Phlatlight modules are used in some of the cheapest disco lights.
Those are maybe reject … used to buy reject high power led for cheap …
TDI = Time Domain Integration ?
I've stripped large projectors, looks similar, wonder if some of those castings are a magnesium alloy?
Can this be used as an astrophotography camera?
Yes, but may need a shutter to avoid light during readout
That looks like a finger-lakes filter wheel, however, for astronomy they'd never use such a large stepper.
PFG is Pulse Flush Gate according to the datasheet.
That LED module says it can go up to 28 amps!!! 21 amps for 100%. You should see what it does at 20 amps!
melts
"We spared no expense!" John Hammond Jurassic Park.
*(that's why this thing costs the same as a 50-seat Greyhound Bus coach!)
That leds look like from led projector
Very interesting. 2x👍
*Summary*
* *0:00** High-End Optics:* The system utilizes heavy, high-quality lenses and mirrors for precise imaging, weighing around 4 kilos each.
* *0:49** Narrow Band Filters:* A filter wheel with five narrow band filters (488, 525, 570, 630, and 700 nm) ensures accurate fluorescence detection and rejection of excitation light.
* *2:01** Customizable Illumination:* The Lumen Dynamics light source offers five individually controllable LED wavelengths (430, 475, 520, 575, 630 nm) with varying power outputs.
* *3:45** TTL Control:* The light source is controlled via a simple TTL interface, enabling easy on/off switching for each LED color.
* *12:55** Sophisticated Camera:* The system includes a 50-megapixel Kodak KAI-50100 CCD camera with a Peltier cooling system for reduced noise.
* *14:54** High-Speed Data Transfer:* The camera features dual analog-to-digital converters to manage the high data throughput of the 50-megapixel sensor.
* *18:11** Possible Issues:* The video creator noted some potential issues with the camera, including image smearing and a limited dynamic range.
* *11:45** Low Runtime:* Internal data suggests the system has seen minimal usage, with only 20 minutes of recorded runtime for the green LED.
* *20:38** Availability on eBay:* Both the illuminator and camera are expected to be listed for sale on eBay.
*In short, this video details a teardown and analysis of a high-end fluorescence imaging system from Fluidigm Polaris, focusing on its illumination source and camera. While potentially powerful, the creator notes some possible issues with the camera and indicates limited prior usage.*
i used gemini 1.5 pro to summarize the transcript
Keep your AI generated garbage to yourself...
A mate with Channel @extractions&ire could use it
The whole sensor module incl. a 5 grand 50mpix sensor for 49 £.. highest bid atm
- Though also a limited CCD sensor, but for the right buyer its a steal at these relative low sums.
-
Architecture Full Frame CCD (Square Pixels)
Total Number of Pixels 8304 (H) × 6220 (V) = 51.6 Mp
Number of Effective Pixels 8208 (H) × 6164 (V) = 50.5 Mp
Number of Active Pixels 8176 (H) × 6132 (V) = 50.1 Mp
Pixel Size 6.0 m (H) × 6.0 m (V)
Active Image Size 49.1 mm (H) × 36.8 mm (V)
61.3 mm (Diagonal),
645 1.1x Optical Format
Aspect Ratio 4:3
Horizontal Outputs 4
Saturation Signal 40.3 ke−
Output Sensitivity 31 V/e−
Quantum Efficiency
KAF−50100−CAA
KAF−50100−AAA
KAF−50100−ABA (with Lens)
22%, 22%, 16% (Peak R, G, B)
25%
62%
Read Noise (f = 18 MHz) 12.5 e−
Dark Signal (T = 60°C) 42 pA/cm2
Dark Current Doubling Temperature 5.7°C
Dynamic Range (f = 18 MHz) 70.2 dB
Estimated Linear Dynamic Range
(f = 18 MHz)
69.3 dB
Charge Transfer Efficiency
Horizontal
Vertical
0.999995
0.999999
Blooming Protection
(4 ms Exposure Time)
800X Saturation Exposure
Maximum Date Rate 18 MHz
Package Ceramic PGA
Cover Glass MAR Coated, 2 Sides or
Clear Glass
--
Features ⤵
• TRUESENSE Transparent Gate Electrode
for High Sensitivity
• Ultra-High Resolution
• Board Dynamic Range
• Low Noise Architecture
• Large Active Imaging Area
Applications
• Digitization
• Mapping/Aerial
• Photography
• Scientific
-
Thx for the tear down Mike, always a joy