Yes. You just need to find a somewhat constant voltage somewhere. Which can be made by the LED if the pass transistor and LED switch places. It'll go away when the LED is off, but that's ok because then you don't need it.
But doesn't that put the source up too high?
Sure, I've been turning this thing off by driving the collector resistor with an HC04, or some such.
don't
Not if it's a depletion mode device. Just make sure not to exceed the gate breakdown, not sure what your max battery voltage is.
But driving a blue LED this way out of a single-cell LiIon is iffy not matter how you twist and turn it. The extra 300mV will help, but ...
[...]
don't
Single cell Li-Ion, ~3.5 - 4.1V
An extra 300mV would easily do it. A Ge transistor would be perfect. ;-)
don't
If 300mV does it then why not spring for those two more reisistahs and offset the base?
If it had to be more precise one could consider controllable shunt references and such but probably that ain't necessary.
don't
Your JFET's a nice touch. If the darn things were repeatable we'd all just do it like this and be done:
Vcc | | V ~~>
--- LED | | | |-' | .--->|-. | | | | | \ | / R1 | \ === / GND | | === GND
This works decently well:
Vcc Vcc -+- -+- | | | V ~~>
| --- LED R1 | | | | | | |/ +----| Q1 | |>. D1 V | --- |
don't
Why not do it right? OK, I forgot, S.E.D is really sci.electronics.dummies :-) ...Jim Thompson
Nice, James, I like it.
How about tossing in a diode-connected transistor for D1? -- Just on the premise you're going to build a million of these and you'd prefer to use the same part as Q1 everywhere and there's no price difference.
---Joel
I thought the object of this thread was to have an LED light at Vcc =
+4V ??Like I said previously: sci.electronics.dummies :-( ...Jim Thompson
the
Yes, I said that in my description.
Cheers, James Arthur
the
Nope.
Keith's not a dummy. He posted a decent circuit. This one stay in regulation even lower. You've misunderstood the object, which is why you suggested "cmos TL431". That doesn't work here--its drop-out voltage is unworkably high.
-- Cheers, James Arthur
don't
And your circuit suggestion is ...?
I don't
much
Posted eons ago... CMOS equivalent of TL431... LMV431 ...Jim Thompson
As a low drop-out current source to run an LED off low voltages? How so? (Not CMOS either, is it?)
-- Cheers, James Arthur
LMV431??
The schematic I see above... please tell me how it works reliably. ...Jim Thompson
Ah, sorry I missed it.
You can remind Jim that you did simulate this at 4V too. :-)
I don't
don't
doesn't
much
Isn't that exactly what I mentioned a few sentences above?
But overkill here and a TL431 is a few cents more expensive than a MMBT3904 :-)
Yep it is, datasheet lies, only the bandgap is parasitic PNP's ;-)
I guess I lost track of which thread was which. I thought this thread was about a battery indicator thresholded for +4V :-(
The mirror I recently observed here changes current, at 3V supply, at
1.27% per 1% supply voltage change. So calling it a regulator is a bit of a stretch.Whatever rings your chime ;-) ...Jim Thompson
...
the
use the
=3D
Right. It won't work in series with an LED when the supply =3D V(LED) +
200mV. That's what we're addressing.
Choose the voltage across R2 to be, say 100mV @ Vcc=3D3.6v, and choose R3 to set the desired LED current. At 4.1v (max Vcc of Keith's LiIon), the voltage across R1 will be 1/6th higher, so output current will be proportionally higher.
That's insignificant, brightness-wise.
If Vcc =3D 3.3v, regulated, and Vf(LED) is 3.1v (a realistic value for many), then the output is regulated and easily controlled.
That's loads better than the first alternative suggested, which was to drive the LED from 3.3v logic using a low-value series resistor. That drifts big-time with temperature, and with Vf(LED) lot variations.
A 2nd alternative was suggested using a series-pass JFET with a current-sense resistor. That lost regulation sooner than this last ckt (above).
These are for Vcc > Vf(LED). Earlier I posted a boosting ckt that'll drive a 3.1v LED from Vcc < 3V.
-- Cheers, James Arthur
Did anyone actually simulate it over a range of voltages? ...Jim Thompson
-- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at
Have something to add? Share your thoughts — no account required.
Ask the community — no account required