EL7900

mA,

the

small

(2.048 V)

that is

any

Hide quoted text -

Once we were discussing food. You said "I don't cook." Later, you talked about cooking and I asked you about the contradiction. Your reply was "I lied."

But why don't you ever post substantive things about electronics? You used to. All you do now is brag and carp.

Post your M=11 collapse simulations, if they actually exist.

Sudden dawning... there's no such thing as an isolated diode on a CMOS process.

So, simplest scenario is photodiode made with its anode as the chip substrate, cathode as N-well. Excessive photo current causes substrate current and, without tracing the layout, who knows what.

N-well to P-source-drain (the PMOS structure used as a photodiode) has even worse possible scenarios... a parasitic PNP (I use them all the time to make bandgaps).

I would guess that's all the possibilities with this part, since Intersil doesn't do complex implanted processes like such folk as XFAB or TSMC.

So find a way to limit your light input, or find another part for the job. ...Jim Thompson

My log-matrix thing has a fatal flaw: the unselected PDs are still working in PV mode. Pity. Phil's architecture, with open-drain switches, works because it really isolates the unselected diodes. Or use one 8-input analog mux per 8 PDs, 4051s maybe. How fast do you need to scan the array?

I don't see an obvious way to do a diode matrix out of the PDs, but I haven't had my quota of coffee and sugar yet.

ohm.

quoted text -

Excellent. You've actually begun to think.

Sudden dawning? I suggested that days ago.

My! My! My! Woweee! I "don't cook", like most males, to avoid kitchen tasks ;-) Though I have, recently, due to my wife's hand surgery, become the "chopping king"... I really have that rolling action with a cleaver down to an art.

So find me in a technical lie. You can't. You _will_, however, find me to be quite harsh when criticizing crap designs... such as yours.

Really! It _would_ be useful if you kept up. As well as we know the model, it doesn't collapse, it "sags".

See my post about photodiodes on this kind of process... with possible parasitic actions that can cause "collapse". ...Jim Thompson

So, there are now "lies" and "technical lies."

And you don't criticize my posts in any rational way, you just whine about them.

You finally figured out how to find M from the data sheet, after I told you the value and you said it couldn't be done. That is a little progress, I guess.

Good, you finally figured that out, too. After someone else posted that little fact.

Oh, please! Why do you need to incessantly be a jerk?

Yes. I expected it to be spelled out in the data sheet. Dream on! Your prodding and snickering _did_ cause me to read the parametric table and back out the ratio from IDD and IOUT. Happy now? I doubt it... someone as ugly-hateful as you can't be a very happy person.

Crap! I did it 3 days ago, but didn't post because I thought I was missing something obvious when I saw only a sag... and "collapse" looked "obvious". (BTW, I swagged M=10, because that's just the sort of thing designers typically choose :-)

With my knowledge of CMOS (been doing designs in CMOS since ~1992, and bipolar I/C's for 30 years before that), I'm trying to imagine a layout scenario that will model collapse. ...Jim Thompson

mA,

V)

is

any

Hide quoted text -

Why must you be an incessant Napoleonic asshole? ...Jim Thompson

Message ID? All I've seen from you is your usual vagueness, and generalities; and "informative" statements like, "Yup, looks like a bug. Too much light breaks it." ...Jim Thompson

Il 02/09/2012 19:27, John Larkin ha scritto:

The minimum scan frequency is about 4 Hz for the whole array/matrix.

:)

Marco

I talk technical and you keep spinning it personal, lame insults without any substantive contribution. That's jerkiness to me. JF and Sloman do the same thing, whining without contributing.

Says the guy who "always gets even", who bragged about getting vengence on somebody after 15 years of brooding.

You can do it by putting a LED in series with each one, but that blows the parts budget. (The LED won't conduct at any voltage the PD can reach.)

You can probably get a factor of 2 by doing the PD equivalent of Charlieplexing: Run the PDs in series-opposing pairs, and use HC4052s to connect all of the pairs to +5 or ground, or leave them open-circuit. (You need to drive the common terminal of each HC4052 to

The reverse-biased PD in each pair will control the total photocurrent--all the forward biased one can do is change the bias voltage on the reverse-biased one by a volt or so, which hardly matters.

Cheers

Phil Hobbs

I mean 74HC4051 (1->8) MUXes. (Details, details. Back to grilling.)

Cheers

Phil

Bwahahahahahaha!

But I'm not ugly-hateful... I'm just patient. _Anyone_ who sets out to specifically do harm to me will be subject to my patience ;-) ...Jim Thompson

Patience usually refers to someone who is slow to anger and ready to forgive, rather than someone who nurses grudges across years. Patience is a good thing, and nursing grudges is a bad thing. (This is applicable to all of us, of course.)

Can we lose this stupid pissing contest and get back to drinking beer? It's Labor Day weekend, guys.

Cheers

Phil Hobbs

I hope that he hears you after telling him 4 times.

?-)