d
Ahh, well I had to look up exactly what a push pull looks like. Mostly we just buy someone else's supply, it's cheaper.
At the moment this is a circuit fragment. Powered from a bench supply. Maybe a phihong 24V switcher.
George H.
d
Ahh, well I had to look up exactly what a push pull looks like. Mostly we just buy someone else's supply, it's cheaper.
At the moment this is a circuit fragment. Powered from a bench supply. Maybe a phihong 24V switcher.
George H.
Hey this is a bit crazy, but what about a box with a power strip inside, for various supplies, and then front panel modules. I'm expecting the user to plug things in so of course I'd have to protect against all possible fu's. the polyfuse-zener looks like the answer. (is there a under voltage, over current fault condition?)
George H.
But, that circuit will actually work as is, when selecting a
3.x+ V LED and proper feed back value R to work with the capacitance sitting at the input of the inverter. C maybe small but still, there is enough to get it oscillation.I would say it was a good example to put some one in the right direction for other applications.
You need to research more before exercising the foot.
Jamie
That's a way to power a blue LED from 3 volts or so.
Really brave folk would eliminate the resistor, too.
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
Using a 1nf cap with Schottky diodes I got good results..
I dropped a diode on the sheet and supplied a model card that would do a blue LED. Vfwd, Ilimit, epsilon, etc, to get a LED that comes close enough. I got ~ 12mA with the on cycle.
I found the selected schottky from the data base, first one, had enough Cj to soften the blow in the pump cycle so that the LED didn't get impacted that heavily with inrush.
Plus if you include some real life "Ron" for the diode in the model card's parm list, it also helps to bring it to reality.
Jamie
Jamie
A good blue LED, nowadays, looks pretty good with 1 mA of average current.
The LED wouldn't mind the current spike.
Given a hex schmitt HC14, you could build a 5-stage ring oscillator, with the extra inverter available to double-up the output drive strength, or to use for something else. At 3 volts, it would oscillate at a few MHz maybe. Then you'd pick the charge-pump cap value to set the LED brightness. Or make a voltage-output charge pump, the original topic.
Hey, an HC14 ring oscillator would make a (not very good) VCO. Or a (not very good) temperature sensor. Or a (pretty good) random number generator.
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
The InGaN greens look pretty good at low currents, too, and easily operated off a 3.3V supply (which surprised me at the time).
I tried an Avago green LED, up in Truckee, where it's really dark at night. After I was dark adapted, I could just barely detect it making light at about
800 pA forward current.Here are some numbers for some Osram LEDS.
so they could be run from 3.3 volts with just a resistor, with some large supply sensitivity. 3 volts or less gets dicey for the blue.
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
Your numbers for the greens are similar to what I found. IIRC, I ended up using a NPN-NFET current source directly from a LiIon battery set at ~5mA (and from a regulated 3.3V supply on wall-powered versions). The LEDs were plenty bright to see in direct sunlight. Like I said, I was surprised I could get away with it but it worked out fine because 5mA was plenty for these LEDs. The 525nm green is good looking, too, but expensive.
You can buy bipolar transistors, like BCX70K, that are beta graded about 2:1. That makes it feasible to make a constant-current source with just one transistor and one resistor, that would in some situations regulate LED current a lot better than just a resistor.
The little HC14 charge pump thing will make the current fairly constant.
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
I don't do beta dependant designs. Too much chance of problems down the road. Decent "logic level" NFETs aren't all that expensive.
George - if you need more current, shift to the 74AC14. They are spec'd at around 25 ma. sink & source. For getting a bit of boost, if memory serves, with an approximately fixed load, a series inductor will work. That was tinkered with but found unneccessary.
Hul
George Herold wrote:
Almost any conceivable bipolar transistor design is beta dependant!
An npn-nfet current source, at least the common one, drops up a couple of volts. If you have that much voltage to spare, just use a resistor.
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
Nonsense, to anyone short of the Slowman class of pendants.
Again, nonsense. Vbe is less than .6V at low currents. The FET simply flatten the I-V curve, and it works rather well.
You don't check the min/max beta specs when you design a transistor circuit? You don't make sure it will work with any transistor in those limits? That's all I'm suggesting.
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
Thanks Hul, I was getting ~5mA from the charge pump thing. That may be enough, but I'll most likely use the LM2662 IC. Re: the inductor... well that didn't work out as well. I was thinking along these lines..
George H.
-- Being a Schmitt-trigger inverter, the presence or absence of input capacitance is irrelevant and the hysteresis built into the input along with the prop delay through the gate will be enough to get the HC14 to oscillate with only the output connected to the input.
-- Larkin back-pedaling??? Congrats, krw, I love it! John Fields
Not a bit. I'm trying to figure out whether he does math or just follows rules.
Because you don't understand it.
Come on, say something on topic once in a while. You're being a squaking old hen again.
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.