Teaching here seemed to always say 1/100th c
Maybe you're thinking of electron drift, which is a good ten thousand times slower still.
AWOL:
Maybe you're thinking of electron drift, which is a good ten thousand times slower still.
Possibly, I went to to study chemistry at university level so yeah. A small misconception only addressed 10 years later....
Now put a photoresistor at the circle center, hooked up as half of a voltage divider. When it receives light, its resistance drops, and you can measure a sudden jump at the midpoint of the voltage divider.
A photo resistor is not going to be able to respond at anything like that speed. You will be lucky if you get 500Hz from one.
Photo diodes on the other hand are much faster but they need amplification and specialized high speed op-amps.
I think your problem is that you do not know enough about electronics to make sensible design decisions. As well as being miss taught or miss remembering lessons.
Well, Grumpy, you certainly live up to your name. I spent a career in the design end of the electronics industry, and when we were working on a tough problem, we "brainstormed" it. That means, toss out your best guess, and then other folks around the table would weigh in with possible changes, or point out parts that might need improvement.
This idea is fundamentally sound, but I don't claim to be an expert on high speed parts. I would expect it to take a lot of development, but it is at least a different approach. As for being "miss taught", I never have had any formal electronics training, but I made it to chief scientist of two tech companies. Perhaps you've been "miss taught" how to spell.
That means, toss out your best guess, and then other folks around the table would weigh in with possible changes,
Or they'd tell you it was hopeless, and you'd move onto something better and more plausible.
Perhaps you've been "miss taught" how to spell.
Perhaps you were never taught that units named after people (other scientists, like Heinrich Hertz) are capitalised, e.g. "100 MHz"
jrdoner:
Well, Grumpy, you certainly live up to your name. I spent a career in the design end of the electronics industry, and when we were working on a tough problem, we "brainstormed" it. That means, toss out your best guess, and then other folks around the table would weigh in with possible changes, or point out parts that might need improvement.This idea is fundamentally sound, but I don't claim to be an expert on high speed parts. I would expect it to take a lot of development, but it is at least a different approach. As for being "miss taught", I never have had any formal electronics training, but I made it to chief scientist of two tech companies. Perhaps you've been "miss taught" how to spell.
@jrdoner, I have to agree that your ideas reveal a poor understanding of electronics. I'm sorry but that is an obvious bare fact. It doesn't diminish your achievements, but it is something you should consider. Perhaps you should undertake some formal studies.
Firstly, I can only say thank you for all the help.
I have taken a lot of this on board and still semi-swaying toward a basic counter. There are fewer parts to have to mangle together, it "may" work and does not cost a lot if it does not!
Indep Factor: Distance of source.
Dependent: Pulse counts.
A 50Mhz pulsed red LED giving a period of 20ns.
100m of fibre (Ir=2ish?) which is an equivalent of maybe 200m of "air" with the light path and the refractive index.
200m/3E8ms-1 = 667ns (looks like I may summon a daemon whilst doing this...) for the light to get through the 100m of fibre cable.
There is a 250MHz capable photodiode:
I wanted to use say a dual 8 bit counter but could only find dual 16 bit counters 16 bit Dual Counter. Sadly, they only resolved to 40MHz where I want a minimum of 50MHz...so 2 8 bit counters instead:
8 bit counters
The counters seem to have a pin that will stop the counting when high. So one counter will reach its maximum value and the carry bit can be use to stop the counting on the second counter. I can then read the difference in counts.
As one count = 20ns, a journey of 667ns = a difference of 33 counts would be expected due to the lag of light through the fibre cable.
Is this a sound plan...or crazy?
Crazy.
What makes you think that those counters work at 50MHz? The data sheet says e maximum frequency is just less than 39 MHz wig a following wind and only 4MHz over the full temperature range.
You underestimate the difficulties of getting counters working at hig speed, it is not so easy as you might hope.
High speed pre scale counters use ECL ( emitter coupled logic ) and that can go at 500MHz but it is a sod to use. I don't know if you can get it these days.
On saying this:
http://www.jensign.com/sol/
Seems pretty much exactly as I want with sensor and source but instead of using an oscilloscope (which I guess I could...) I can use the counters instead.
Grumpy_Mike:
Crazy.
What makes you think that those counters work at 50MHz? The data sheet says e maximum frequency is just less than 39 MHz wig a following wind and only 4MHz over the full temperature range.You underestimate the difficulties of getting counters working at hig speed, it is not so easy as you might hope.
High speed pre scale counters use ECL ( emitter coupled logic ) and that can go at 500MHz but it is a sod to use. I don't know if you can get it these days.
OK, maybe this is where my lack of qualification in electronics comes in...
The data sheet says at 6.0V 25oC, a frequency of 61MHz is "typical" and 39MHz is the minimum. Am I reading it incorrectly? Does it mean as minimum it is 39MHz and the 61MHz is like the absolute best with the best clock source/most stable temp/no noise/no RF noise etc?
Grumpy_Mike:
You underestimate the difficulties of getting counters working at hig speed, it is not so easy as you might hope.
High speed pre scale counters use ECL ( emitter coupled logic ) and that can go at 500MHz but it is a sod to use. I don't know if you can get it these days.
CMOS has eclipsed ECL for high speed logic. There is a reason why you can't get ECL these days.
charliesixpack:
CMOS has eclipsed ECL for high speed logic.
There's a lot of advantages bundled into chip scale integration there too.
Does it mean as minimum it is 39MHz
Yes.
and the 61MHz is like the absolute best with the best clock source/most stable temp/no noise/no RF noise etc?
No.
The 61MHz is what you might get, the 39MHz is what you should rely on. Also look at the temperature range, that was only for a narrow band.
In other words you are working on the edge of what is possible with this device. So you can not say that any one device you will buy will work at 61MHz, but if you buy a big bag then typically many will work at the faster speed but if any don't work, as long as they work at 39MHz then tough luck.
The 39MHz assumes "best clock source/most stable temp/no noise/no RF noise etc"
It is a bit like over clocking a processor, some seem to work other don't and the ones that seem to work could stop working like this at any time.
Grumpy_Mike:
Yes.
No.
The 61MHz is what you might get, the 39MHz is what you should rely on. Also look at the temperature range, that was only for a narrow band.In other words you are working on the edge of what is possible with this device. So you can not say that any one device you will buy will work at 61MHz, but if you buy a big bag then typically many will work at the faster speed but if any don't work, as long as they work at 39MHz then tough luck.
The 39MHz assumes "best clock source/most stable temp/no noise/no RF noise etc"
It is a bit like over clocking a processor, some seem to work other don't and the ones that seem to work could stop working like this at any time.
Oh god. Now datasheets get more confusing the more I use them...
So what determines a minimum of 39MHz? Is it within one standard deviation with a mean of 39MHz for their brand of the chip?
Is the 61MHz a 2 sigma possibility from the 39MHz? They just seem to be like "random" values they have selected. If the reliability is as bad as "it may work at 39MHz-61MHz" then it is as much use of a quote as "A car is always red".
How are these things determined in essence? Do I get 3 chips and try them all out? Do I expect with a 95% CL I get 39MHz and like a 0.3% chance of getting one at 61MHz?
So what determines a minimum of 39MHz?
It is what they can manufacturer most of the time.
They just seem to be like "random" values they have selected.
No it is based on their yield information from what they can make from their fabrication plant.
A good design will always take the minim value and not exceed it. A medium design will take the typical value and test the product 100% to the limits it is using. A bad design will hope it will work and not test it. That is what many China manufacturers do.
Do I get 3 chips and try them all out?
You could, but remember that if it doesn't work then it might not be the chip it might be what your are doing to the chip. It is yet another unknown and in electronics there are too many unknowns already so adding an other one is just an unnecessary complication.
Do I expect with a 95% CL I get 39MHz and like a 0.3% chance of getting one at 61MHz?
There is no way of telling anything like this.
Johnny010:
Oh god. Now datasheets get more confusing the more I use them...So what determines a minimum of 39MHz? Is it within one standard deviation with a mean of 39MHz for their brand of the chip?
Nope. Suppose you're making a million chips. You design them to work at the fastest speed possible (60MHz). When they are produced, you test every one. Most are able to operate at 39MHz and above. They pass the test. The remainder get binned (or recycled). How will you ever know if the chips you buy will be able to operate at 60MHz? Even the manufacturers don't know which of their chips will do that, because they don't test for it. Of course, some will operate at that speed, maybe as low as 1% of the passes. Who knows?