Phototransistor fed to a transimpedance amplifier

Yup... I'm just back from vacation, where I took Mark Johnson's "Photodetection and Measurement" to the beach for reading.

He spends a chapter on making measurements in industrial environs. (The book would be a good predecessor for Phil's optics text...)

George H.

Yup, good book. I gave it a nice review on Amazon about a dozen years ago. Mark and I corresponded for a couple of years before the book came out, and have occasionally since.

He's particularly good at hands-on demos (the 'Try It' sections) and on demystifying lock-in amplifiers, which as he shows work exactly like narrow bandpass filters followed by downconverters.

I just recently received a paper to review from SPIE's Optical Engineering journal whose main idea was one that I recognized from Johnson (it's on P. 185). I doubt the authors were expecting anyone to notice. (Peer review is such a joke nowadays.)

Cheers

Phil Hobbs

Cheers

Phil Hobbs

You need a differential TIA with good common mode rejection. And to get good common mode the two diff inputs need to present the same Z to ground. (unlike the typical diff amp)

Goggle common mode rejection

The circuit is so slow that there's no reason I can see not to use the phototransistor barefoot with a load resistor, and bypass the receiving end like mad. If there's some really gross amount of RF around, 1 nF from B-C right down at the phototransistor will help.

Cheers

Phil Hobbs

Hmm, do you have his email address... I wanted to send him a thanks for the book. And also talk a little about 'scope FFT's and Zener diode noise sources... (I made notes in the book margins.) Perhaps send it to me off-line.

George H.

Sent. He's also on LinkedIn.

Cheers

Phil Hobbs

Hey but dead chickens are great, I keep one by my scope at all times!

Tim

You forgot the most important "parts"...

Eye of newt, and toe of frog, Wool of bat, and tongue of dog, Adder's fork, and blind-worm's sting, Lizard's leg, and howlet's wing,-- For a charm of powerful trouble, Like a hell-broth boil and bubble. ...Jim Thompson

Yeah, but yours is deep fried, which doesn't count. ;)

Cheers

Phil Hobbs

Macbeth (IV, i, 14-15)

Shakespeare said it first. Jim may have learned it as his mother's knee (or some other low joint) but I suspect plagiarism. MIT graduates aren't famous for plagiarism but it does happen.

Flatman and his side-kick Ribbon?

What can I say? I'm a very philosophical monk. (A deep friar.)

Tim

Just say no to analog light sensors on long cables. Go to mouser, search for this:

taos Optical Detectors and Sensors

Good Luck.

A phototransistor you can bypass. You really think I2C over 4 metres of cable is an improvement? You're definitely a sport. ;)

Cheers

Phil Hobbs

I was thinking of something more like this:

It has a chance of fitting into the available space. The frequency output is probably a better choice for running with the motor cables than an analog signal. Also, you get a free spectrometer, so you can filter by color.

Who was it who said never, never put a photo detector on the end of a cable? :)

Same guy who said "never use a phototransistor for anything". ;)

But the OP's application seems to be very low speed, so a barefoot phototransistor plus heavy bypassing should work fine. A 100-ms time constant gets rid of a lot of nasties.

Cheers

Phil Hobbs

I know. At first I wondered if you had been kidnapped and were trying to embed secret messages for help. I'm glad you're OK.

If a lot of phototransistors were all in a pile, it might make a good paperweight. Maybe pour over with clear epoxy?

Hmmmm, maybe. Even running 110V for instrument power through trays with motor power is chancy, much less small signals. I got a paid trip to the Aleutians once because a PE broke that rule.

Hey OP, u some place interesting? U use cable, I come fix.

ChesterW

Personally I'd probably burn them for winter warmth. ;)

Yeah, VFDs are seriously bad neighbours. You can get millivolts of crud across solid metal frames that are probably down in the microohms. Circulating current of some amperes, with millivolt drops, is what make ground loops so seemingly mysterious. Jiggling a wire changes the contact resistance and effectively makes or breaks the ground loop. Nickel-plated connectors make this much worse, because they form Ni-NiO-Ni tunnel junctions.

Did you get to visit Spruce Island? I hope to do that one day. I've only been as far as Homer. Terrific place in the summer.

It's supposed to be 114 in Riyadh tomorrow. ;)

Cheers

Phil Hobbs

This is the VFD for a vent fan on the roof. It was making 20 volt spikes on the other side of the building. It was a nuisance to filter.