Oh, your sim is crap. Look at the currents through D2 and D3!
You added R3 to the second circuit to make up for the absurd schmitt model you're using.
If you try hard enough, you can break any circuit and reject any idea. Keep it up.
John
This is nice:
+3.3 | | led | | c +3.3-------R-------b BCX70K e | | | gndR is 330K maybe. Current will be around 5 mA down to Vcc-Vled of maybe
0.2 volts. The BCXs are beta-binned, plenty tight enough for an LED.John
-- It's not my sim that's crap, it's your circuit that's allowing that to happen.
Hang a 0.1 ohm resistor from the schmitt output to ground. Measure the current. Report back.
That's no HC14!
John
Sure is temperature stable, too...
:-P ...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 http://www.analog-innovations.com | 1962 |
The only thing bipartisan in this country is hypocrisy
Surf on "LM3909 replacement with low power consumption" ...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 http://www.analog-innovations.com | 1962 |
The only thing bipartisan in this country is hypocrisy
--- And there's no need for it to be, since what we're interested in finding out about your circuit is, initially, how it behaves using a perfect driver. (well, a driver with with a 1 ohm source resistance, which is what LTspice's gates default to.)
With that in mind it's clear that even with a perfect driver, your circuit, which pushes 150mA spikes which decay to essentially 0mA in
4ns every 10 microseconds or so through the LED, won't work.More's the pity for a poor little HC14 running at 3.3V, which can only source or sink currents on the order of 10mA or so before the Rds(on) of the output transistors starts dropping enough voltage to starve the LED.
The conclusion then is, I'm pretty sure, that since you specified a Schmitt trigger inverter for your driver which needs to be made out of unobtanium, your idea/design/sketch/scribble/whatever, as drawn, is dead in the water.
Oh, and about getting rid of the timing resistor to speed things up?
If you fix the circuit, you might want to talk to the FCC before you start selling anything...
Just an idea...
JF
And it has picosecond rise and fall times. Which is why there are insane currents *in both directions* in the diode and the LED.
It actually does work with a perfect driver, as your sim shows. It works better with a real-world CMOS part and properly chosen values. You don't believe in charge pumps?
You also boogered the schmitt thresholds to totally unrealistic values, for some reason. Probably to make the second circuit work.
You were just complaining that the circuit pushes too much current into the LED, now you are complaining that it pushes too little.
The output impedance of an HC14 is a nice match to the LED. The average LED current depends on the switching frequency and the value of the cap. I = CVF and all that. That can be *designed* as needed. Or broken, if you prefer.
But ancient HC, and especially modern Tiny cmos parts, can drive a lot more than 10 mA at 3.3 volts.
You picked a perfectly stupid Spice model for the schmitt so that you could make it not work to your satisfaction. Do you specialize in making things not work?
I said it would scream. But there are lots of digital circuits around, on BIG boards, that run at hundreds of MHz. We do that all the time.
Four dacs, clocked at 128 MHz:
ftp://jjlarkin.lmi.net/DSC01786.JPG
Scary stuff. Being a 555 guy, you probably don't deal with MHz very often.
My circuit is similar to Spehro's. Attack him for a while.
At least you got the schmitt to oscillate, as JT said it wouldn't. Attack him, too. Canes and walkers at 20 paces maybe.
John
As shown.
"Engineered" all the way ?:-) This Yo-Yo has a history of ba-a-a-ad circuit posts. ...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 http://www.analog-innovations.com | 1962 |
The only thing bipartisan in this country is hypocrisy
-- That's nice??? It seems kinda fugly to me... It's just a plain old common emitter, sorted for beta and forced into taking center stage, while Hfe(t) and Vce(sat) are really running the show. On one hand it might stop the LED from glowing when Vbe = 0V but, on the other, what is it you don't understand about: "I need to drive a blue led from 3.3V. Most of the SMD blue leds I find have a Vf equal to or greater than 3.3V. I am curious how others are dealing with this. Thanks Rich" Clearly, by interposing Vce(sat) between the supply and the LED you've reduced the current available to the LED and dimmed its output in a clumsy way. Can you dig yourself out of this pit into which you've fallen with other than invective?
You got it!
Read my reply to him. I noted that most blue LEDs actually need less than 3.3 volts, and that if his really do need more, use a charge pump. If his LED does work on 3.1 or less, my little transistor thing delivers a decently constant current. If they need much less, a resistor will do. I think I said that somewhere too.
If I said that iron is a metal, you'd crank up Spice and prove it wrong.
Since you have proven by simulation that charge pumps don't work, I'm in big trouble. So is Maxim.
John
IWith that in mind it's clear that even with a perfect driver, your
-- Nope. Unfortunately, you don't have a perfect driver, so your scheme is bogus, from the start.
You're a very bad circuit designer.
John
[snip]
Speaking of yourself ?:-)
You're throwing some _really_ bad shit on the table, and it's going to bounce back in your face. ...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
And you keep counting on the general amateur level of lurkers here to let you get by with it.
I urge everyone to LTspice anything that Larkin throws out here, particularly taking note of unrealistic clock speeds and resulting current levels. ...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 http://www.analog-innovations.com | 1962 |
The only thing bipartisan in this country is hypocrisy
Looking over the circuit again I don't think that LTSpice inverter models realistic output impedance. I hung a capacitor off the input to model the gate capacitance, but I forgot to put one on the output to substitute for the output impedance. It still seems to work fine with a
100 ohm resistor on the output. I put the 220 ohm resistor there because I wanted to see what output voltage it was creating for different component values, without it there the voltage just drops to the forward voltage of the LED.Has anyone thought to breadboard the circuit up? Surely that would settle the dispute. I don't think I have any modern inverters lying around, but I do have blue LEDs and some 74LS14s somewhere!
It has zero rise and fall time and 1 ohm impedance. It will source or sink 3 amps. And the schmitt thresholds are silly. That's why it's pounding so much current into all the junctions in both directions.
I wish I could buy a gate like that!
I hung a capacitor off the input to
And LS will need a low value resistor to oscillate, ballpark 1K, because of its DC input current. So to get a sensibly low oscillation frequency you'd need a real capacitor, not just Cin of the gate. A CMOS schmitt will oscillate with a 1M resistor, or probably with a 1G resistor.
John
On Thu, 27 May 2010 17:20:49 -0700, John Larkin wrote:
--- Good point; here it is with real-world rise and fall times right out of Philips' Data Handbook IC06:
Version 4 SHEET 1 880 680 WIRE -128 16 -528 16 WIRE 336 16 -128 16 WIRE 336 32 336 16 WIRE -128 64 -128 16 WIRE 128 128 80 128 WIRE 224 128 192 128 WIRE 240 128 224 128 WIRE 336 128 336 112 WIRE 336 128 304 128 WIRE 336 160 336 128 WIRE -528 176 -528 16 WIRE -336 176 -384 176 WIRE -240 176 -272 176 WIRE -224 176 -240 176 WIRE -128 176 -128 128 WIRE -128 176 -160 176 WIRE 80 192 80 128 WIRE 112 192 80 192 WIRE 224 192 224 128 WIRE 224 192 192 192 WIRE -128 224 -128 176 WIRE -384 256 -384 176 WIRE -352 256 -384 256 WIRE -240 256 -240 176 WIRE -240 256 -272 256 WIRE 80 256 80 192 WIRE 336 272 336 240 WIRE -384 288 -384 256 WIRE -528 368 -528 256 WIRE -384 368 -384 352 WIRE -384 368 -528 368 WIRE -128 368 -128 288 WIRE -128 368 -384 368 WIRE 80 368 80 320 WIRE 80 368 -128 368 WIRE 336 368 336 336 WIRE 336 368 80 368 WIRE -528 416 -528 368 FLAG -528 416 0 SYMBOL res 208 176 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R1 SYMATTR Value 1MEG SYMBOL cap 304 112 R90 WINDOW 0 0 32 VBottom 0 WINDOW 3 40 29 VTop 0 SYMATTR InstName C1 SYMATTR Value 10n SYMBOL res 320 144 R0 SYMATTR InstName R3 SYMATTR Value 220 SYMBOL LED 320 272 R0 SYMATTR InstName D1 SYMATTR Value NSCW100 SYMATTR Description Diode SYMATTR Type diode SYMBOL res 320 16 R0 SYMATTR InstName R2 SYMATTR Value 1k SYMBOL cap 64 256 R0 SYMATTR InstName C2 SYMATTR Value 5p SYMBOL cap -160 160 R90 WINDOW 0 0 32 VBottom 0 WINDOW 3 32 32 VTop 0 SYMATTR InstName C3 SYMATTR Value 10n SYMBOL voltage -528 160 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V2 SYMATTR Value 3.3 SYMBOL res -256 240 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R4 SYMATTR Value 1e6 SYMBOL cap -400 288 R0 SYMATTR InstName C4 SYMATTR Value 5p SYMBOL LED -144 224 R0 SYMATTR InstName D2 SYMATTR Value NSCW100 SYMATTR Description Diode SYMATTR Type diode SYMBOL Digital\\schmtinv -336 112 R0 SYMATTR InstName A2 SYMATTR SpiceLine trise 100e-9 tfall 100e-9 SYMATTR Value2 Vhigh=3.3 Vt=1.65 Vh=0.5 SYMBOL diode -144 64 R0 SYMATTR InstName D3 SYMATTR Value 1N4148 SYMBOL Digital\\schmtinv 128 64 R0 SYMATTR InstName A1 SYMATTR SpiceLine trise 100e-9 tfall 100e-9 SYMATTR Value2 Vhigh=3.3 Vt=1.65 Vh=0.5 TEXT -520 392 Left 0 !.tran 0 10e-5 1e-6
Happy now?
The currents are _still_ insane.
---
--- If the criterion determining "works" is whether the LED will emit light which can be detected by the human eye in a casual manner, then I submit that 4ns wide pulses occurring at a rep rate of 100kHz won't quite fill the bill.
---
--- Neither of which you addressed earlier, with your cockamamie circuit, while now you're pontificating as if you did.
---
--- Idiot.
What's that supposed to do, put me on the defensive?
---
--- I just copied Bitrex's part without looking at the values, but changed them, above, to half of Vcc +/- 0.5V, which is more realistic, and it still oscillates.
The fact of the matter is, dolt, you can make the switching points whatever you like and the thing will still oscillate.
---
--- Hmm...
You really _are_ rather thicker than I thought, since one example was with a perfect driver and the other with real-world parts.
However, in either case, your circuit won't work.
---
--- That depends on how much current the LED needs, since the higher it goes, the lower the available output voltage from the gate.
---
--- More platitudes...
---
--- Blah, blah, blah.
---
--- Nope, I just used what was there.
---
--- Nope, I specialize in pointing out things that don't work.
Like, for instance, latching relays with infinite gain.
---
--- Snipped ad.
---
--- Scary for you maybe, but you have no idea what I'm up to.
---
--- Why should I?
He's not an asshole.
---
--- He was right, too, since your initial "offering" was missing a bubble.
---
--- BTDTGTTS
---
Canes and walkers at 20 paces maybe.
---
And you're getting younger???
You still have a silly Schmitt model. It will put out 3 amps, not realistic for a cmos logic part. That's why the peak currents are high.
But as long as the gate can make full swing, the average LED current depends on the frequency and the cap value. And every version you posted, including this one, does push a reasonable abount of average forward current into the LED.
I should hope it works. It's a simple charge pump. Your latest example puts about 5.5 mA average forward current into the LED. You have proved that my charge pump works, even with the unrealistic schmitt model.
Average current makes light. Eyeballs respond to average light. Are you suggesting that X microwatts of light are visible if DC, but invisible if the same average amount of photons arrive in bunches?
Those big roadside billboards use multiplexed LEDs, each LED operating at a low duty cycle. These are invisible to you?
If you claim charge pumps can't work, you've done that to yourself.
If you can't come up with values to make this charge pump work nicely, you've compounded the damage. You so badly want it not to work that you've stopped thinking like an engineer. That's a really bad sign, letting your emotions suspend your ability to design. If you wanted it to work, you'd let it work.
Bitrex's circuit likes an asymmetric duty cycle, which is maybe why he has the thresholds where they are. The asymmetry struck me immediately, because an HC gate oscillator would be nearly symmetric.
In all the cases you've simulated, it does work. So many people do things like this that it's hardly "my" circuit. It was done with tubes before I was born.
John
I'm still new to LTSpice, I was a little confused about how the model sets the thresholds and hysteresis. In datasheets the hysteresis is defined as the difference between the high transition point and the low transition point, in the model Vh seems to be the difference between the high and low transition points and the midpoint, which is defined by the Vt parameter. I think I've rectified most of my errors now...:)
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.