[DIYbio] Re: qPCR fluorescence detection dynamic range

The model piece that I've been using is the AMS linear array. http://www.mouser.com/ProductDetail/ams/TSL1402R/?qs=sKasJQfA%252bi5t6//qMqmpuA== It's a 16 dollar chip that if you spend some time actually playing with it, youll begin to realize that many of the arguments being made here are moot, or mootable.

By controlling the signalling to the chip, you can control the integration time- the amount of time the chip spends "counting photons". Similar (or exactly the same thing) to exposure on a camera, you control the mantissa at the upper end of the range by reducing the amount of time the capacitors have to store the charge, the electrons being released by photons. If you need to resolve at the lower end, you can increase the exposure time to capture more events. 

To the point of "physics intensive"- you waste more time dealing with people who "cant do math" than just doing the math. What we need to do is start breaking apart the jargon. It annoys me when people talk about 16 bit ADC on a 8 bit camera. There's no such thing as an 8 bit camera- there's just capacitors storing charge off a diode, analog doesnt have bits, thats why it is analog. If you're hooking the ADC up right, its the physics of the underlying chip, the diode and capacitor, that really matters: the choice of wavelength (green is usually better), the tolerance and noise and things like "dark current" that dictate how well a 8-, 10-, 24-, 1000- bit chip are going to read that charge.

256 px doesn't seem like a lot, but it's all about controlling not just things like slit width, but also integration time and averaging, as well as all of the components of Beers law. A cheap ADC (Arduino supplies 10 bit) can cover way more range if you have control over exposure time, and attempt to adjust or autorange as best as possible. It would seem that for DIY qPCR, sinking all of your money into pricey detectors is wasted if your optics are crappy, and that sample averaging and exposure control can squeeze whatever resolution you need provided that you have decent control over the optics of the system.

I think part of the problem I'm having engaging the DIY community is that there's a willingness to salvage parts but not as much attempt to document and share what people learn along the way. Taking apart $$$k equipment is fun and all, but it seems like we're losing the basic skills that were state of the art 20 years ago, like the power of what you can do with a single, crappy cheap ass photodiode or photomultiplier. Yes, "its been done", and it's not what the cool kids at Illumina or BioRad are doing, but often simple can force you to think of new ways of looking at the problem. There's no reason why these kind of experiments aren't open sourced and repeated to death so that everyone, no matter what their experience level, can begin to communicate and share with others.

lets teach the physics, the math, the basic science, and not just gloss over it. do, don't debate. Grab an arduino and a photodiode and capacitor and do the experiment. mess around!

-matt