How start a transistor at 3V when signal is very high impedance?

There's may be an easier way, but you could put a 2.5V reference chip (LM385-2.5) between the emitter and ground. The transistor would start to turn on at about 3.1V wrt ground. You might need to have a resistor from plus supply to the reference chip to get it stabilized.

-- John

Interesting idea. I hadn't thought of anything between emitter and ground. I wonder if I could use a diode chain there instead - 3 diodes to drop around 2V, perhaps, or four to drop about 2.5V. I have the sneaking suspicion this is not the right way to do it but it might work and still be a low impedance path to ground. How does that sound?

You could, but the turnon will be more gradual. For either method, the impedance to ground will decrease as the input increases, so it may still load your input some.

-- John

If it's "very high impedance" as the subject says, use a comparator. Especially if you care about the precision of "around 3V".

--
Ben Jackson AD7GD

http://www.ben.com/

If you need to turn on a transistor at some arbitrary voltage, then a comparator (with hysteresis in this case) would do the job. Comparators are cheap (well under $1) and would do the task you need.

Cheers

PeteS

use a PNP transitor that pulls down a resistor from the + side down to around 0.6 or so.. your 3 volt input can then reverse bias the PNP which will allow the resistor to then drive the base with more voltage.

just a thought.

--
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5

[OP] Ok I have been looking at comparators which I was not aware of and they seem ideal. For the positive voltage of c. +3v I think even I could get it to work though I may have some issues with the potential input range to +15v, possibly more, running with a 9v battery.

The negative comparator is more of an issue. It looks as though I need a -3v reference unless a comparator will compare its non-inverting and inverting inputs only with each other and not in reference to its own ground. I can't quite make that out. Is that what they do? Or are there some that work in that way?

BTW, a quad comparator looks good, especially as I've realised I will need one for each of the 16 LEDs. Four chips is reasonable.

If you mean the inputs can go to 15V while the comparator is running on

9V, that could be an issue. You can get around it by including a largish resistor (say, 1 megohm) in series with the input, and using diodes to clamp the input to a safe voltage. So if your reference input is at a stiff 3V, you could connect back to back diodes from the other input to it, and it would work for comparison to 3V. The diodes and resistor would have no effect until the input went over 3.6 V or less then -3.6V.

The two inputs are compared with each other. If the + input is higher (more positive) than the - input, the output goes all the way plus. If the - input is higher than the + input, the output goes all the way minus. Its just a differential amplifier with very high gain.

-- John

Jamie wrote: snip

I've tried a few times to understand this but I don't follow? I would still need some way to generate the 3v so the NPN doesn't turn on until that has been exceeded. Not a major problem now, if using an op-amp or a comparator rather than just an NPN.

yes, put a 2.4V drop between the input and the base.

for high impedance use a comparitor eg:

+12V | ----o out |\\ 1K / in----|+\\ | |/ | >---+--| +3V---|-/ |\\| |/ ~\\ | ---+--- 0V Bye. Jasen

using an op-amp is also a good way,

--
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5