EL7900

That figures.. Your common phrase must be:

"I am never wrong, once I thought I was, but that turned out to be a mistake"

With all your retractable statements, you still managed to make comment on how you correct people, JL in this case.

Who corrects you? Oh wait, I forgot, you're never wrong..

Jamie

rk.

e a

You don't pay much attention, do you. There are things that I knew that turned out not be so, and some of them get exposed here.

It doesn't happen all that often, but it has certainly happened.

It varies. I'm damned if can remember who did it most recently - I had an idea that it was Tim Williams but the search engine doesn't throw up anything.

-- Bill Sloman, Nijmegen

He is AlwaysWrong, you know.

No, I am not, and you are doing it again.

Il 05/09/2012 00:03, John Fields ha scritto:

You drew a led coupled to the photodiode. Do you recommend any kind of led (white, green, etc...)? What's the difference between using a led instead a lamp or sunlight?

Marco

Il 05/09/2012 11:26, John Fields ha scritto:

Well, I have only standard green leds (low-power) and they can't drive the device properly.

So I used several light sources: white led, incandescent lamp and halogen lamp. With a supply of +5V the maximum output current measured was about 4 mA no matter how close the light source was. Unfortunately today I can access only to a DMM with 1 mA of resolution.

With a supply of 3.3V I read max 3 mA.

I didn't see any collapse. I'm curious to redo the test with direct sunlight - when possible.

Marco

Allow me.

*Whoosh!*

--- Good!

That means that with only the low resistance of the milliammeter in the circuit, the voltage dropped across it doesn't run into the Vcc-0.3V compliance limit of the device.

Take a look at Table 1 of the data sheet and you'll notice that for any illumination level there's a maximum value of load resistance allowed, the reason being that if the resistance goes higher, then the output voltage will increase to the point where it exceeds the voltage compliance limit for the device.

You can prove whether that's true or not by doing this test:

Vcc / .+5-------+--------+ . | | . [R] | Vout . | +--+--+ / . [LED]-->| Is|---+----+----+ . | |_ | |5k | | . +-----|E | [POT]-----+--------+------+---------+

Start with the pot at minimum resistance and illuminate the device to get some current - say 3mA - out.

Then slowly rotate the pot and eventually you should get to the point where the collapse you noticed in your circuit is duplicated.

Then, subtract Vout from Vcc, and the difference should be less than allowed by the data sheet specs.

If the collapse happens, then what it means is that your load resistance is too high for the level of illumination being used and you should either decrease the illumination or the load resistance or increase Vcc, or both.

---

--- I'm all ears!

-- JF

:) Jamie

Oh boy!

Ahhh... Shaddup!

you

put

ing

his

work.

s.

of

be a

It seems to have been Phil Hobbs in the thread "LTC Switcher CAD3 ?" on the 4th June.

I was talking about nanohenries per root turn, which is a mistake I'd been making for years, and I didn't immediately realise that Phil was right when he told me that it should have been Henries per turn squared. It was three in the morning, my time, and I should have gone to bed rather than responding without thinking hard enough about what I was posting.

It was embarrassingly obvious that he was right when I looked at the thread the following morning, and I did post an acknowledgement immediately.

You should have remembered. You chimed in almost immediately, and - as usual - said absolutely nothing that was interesting or useful.

-- Bill Sloman, Nijmegen

You're no better a poet than you are a circuit designer.

-1

That's unkind. He's an uninspired circuit designer - from what I've seen - but it's not as bad as his poetry.

Good theme though. Someone with talent might make something of it.

-- Bill Sloman, Nijmegen

Make that a long long int.

Jamie