FLASHING BULB CIRCUIT WANTED.....

ash

The usual source of "random" signals is a pseudo-random binary sequency generator.

The sequence of "0" and "1" states it produces isn't actually random, but a long enough shift register won't repeat itself for quite a while and is random enough for your kind of application,

Horowitz and Hill's "Mastering the Art of Electronics" has a useful section on the subject in chapter 9.32.

These days a professiona engineer would realise the shift register and the exclusive-OR gate in a programmable logic array or a single-chip microprocessor, but you can still buy cheap CMOS shift registers and OR-gates, and use the shift register's parallel outputs to drive the gates of 25 cheap N-channel MOSFET transistors to control your lamps.

-- Bill Sloman, Nijmegen

ash

A little off the beaten track, but it might work. (?)

Get a box and attach little strips of metal inside so that the edges line up close to each other. Each pair of strips will form one switch. The gap between the metal strips should be small, maybe 0.1 inches apart. Connect it all up.

Then fill the box with ball bearings and shake. Adjust amount and size of ball bearings until you get the effect you want.

MacGuyver up some way to shake the box and you're done!! :)

--- There are at least three problems with that approach, the first being that it won't at all simulate incandescent lamps with mechanical flashers, because the flashers on the lamps all had a more-or-less uniform period, but the differences between the lamps kept them from syncing up, giving the appearance of random flashing in the array.

The second problem is that if the lamps are wired to follow each other like the shift register outputs, then the appearance will be that of a "chaser", with the pattern of on and off lamps rolling from one end to the other. That can be cured by wiring the lamps haphazardly as to which output they're connected to, but the third problem, which is that no matter how long or short the lamps' on or off times might be, when a change occurs it will be on a clock edge, which will surely kill the illusion of randomness.

Here's how I'd do it:

. VCC . | . +-----------+ . +-----------|---------+ | . | |8 | | . [Rt] +----+----+ | | .VC>-----+---+ | |_ Vcc | | | . | | C--+ O|D VC| | | . | [LED]--> B | _| | | . | | E--+----6-|TH R|O-4---+ . | | OPTO 1 | |__ | | | . | | +---2-O|TR OUT|-3--+-|--->OUT1 . | [R] | | GND | | . | | [Ct] +----+----+ [0.1µF] . | | | U1 7555 |1 | .GND>----|---+-----------+-----------+-----------+ . | . . . . . . . | VCC . | | . | +-----------+ . | +-----------|---------+ | . | | |8 | | . | [Rt] +----+----+ | | . +---+ | |_ Vcc | | | . |A C--+ O|D VC| | | . [LED]--> B | _| | | . | E--+----6-|TH R|O-4---+ . | OPTO 25 | |__ | | | . | +---2-O|TR OUT|-3--+-|--->OUT25 . [R] | | GND | | . | [Ct] +----+----+ [0.1µF] . | | U25 7555 |1 | .GND>--------+-----------+-----------+-----------+ Here's how it works:.

There are 25, 7555 astables with equal values of Rt and Ct, so the different component tolerances will cause the astables to oscillate at approximately (+/- 10-20%?) the same frequency so the lamps will always turn on and off out of sync.

For a speed control, the collector-to-emitter junction of a phototransistor is connected between the 7555 output and the anode of the opto's LED, so as VC is varied the current through the 25 LEDs will also vary, effectively changing the value of the 25 charging resistors, causing the frequency of oscillation to change.

But, since you hate 555s so much, an even simpler way to do it, which would reduce the package count significantly, would be to use logic gates instead of 555's, like this:

. VCC . |K . [DIODE] . | . +--------+-------+---[Ct]--+ . | |A | | . | [DIODE] | | . | | | | .VC>--+---|---+ GND [Rt] | . | | | | | . | | [R] | | . | | | C---+ | . | | [LED]--> B | . | | | E---+ | .GND--|-+-|---+ | | . | | | | | . | | | | \\ | | \\ | . | | +---------| >O--+--| >O--+ . | | | / | / . | | . | | . . . VCC . . . |K . . . [DIODE] . | | | . | | +--------+-------+---[Ct]--+ . | | | |A | | . | | | [DIODE] | | . | | | | | | .VC>--+-|-|---+ GND [Rt] | . | | | | | . | | [R] | | . | | | C---+ | . | | [LED]--> B | . | | | E---+ | . +-|---+ | | . | | | . | | \\ | | \\ | . +---------| >O--+--| >O--+ . | / | /

"John Fields" schreef in bericht news: snipped-for-privacy@4ax.com...

The 555 solution looks promessing. But you may need to take measures to prevent the individual circuits from "seeing" each other. I once experienced that similar circuits synchronized after some time. Never had the change to look after the cause but I guess they influenced each other via the power lines. For the same reason the gates-solution may not randomize.

petrus bitbyter

flash

ust

Steve doesn't seem to want to simulate lamps with mechanical flashers

- he's merely asking for something similar.

This doesn't seem to cut into the sales of LED-based flashing lights

Sonce the 555 doesn't have a lot of common mode rejection, this is unlikely to be true. I suppose that anybody silly enough to propose wiring up 25 555-based drivers has to be silly enough to be unaware that buying twenty-five identical resistora and capacitors is likely to get you parts from the same manufacturing batch, which will be much closer in value that the specified tolerance, so there might not not be a lot of unit to unit variation between the array of 555-base flashers to prevent the switching spikes from one part from triggering all the other parts that are close to switching, and eventually forcing most of the flashers to cycle in concert.

The kind of elaboration that often goes wrong ...

You really do want Steve to do a lot more soldering that he has to, don't you.

Are there an endless supply of wet-back wiremen in Texas who will work for peanuts?

-- Bill Sloman, Nijmegen

PLA's? uP's??? That's like using a cruise missile to kill a mouse. Google "relaxation oscillator". P.U.T.'s are still available from ON Semi (and maybe others) for about a dime ea. Freq can be varied by varying the charging resistor. Add an NPN to the output to drive the load.

What is your solution Bill?

Ed

None. He's all critique and no solution, just like Ms Prissy ;-) ...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

Interesting problem. Kudos to mpm (alias Rube Goldberg) for the ball bearing solution. LOL

Don't they still make LED's with blinking circuit built in? Do those have good common node rejection? Do they end up synching up with each other?

Could the LED's with blinking circuit inside be used to drive FET driver transisters?

I think it's a given that, unless you have phenomenal VDD filtering, relaxation oscillators tend to at least "semi-sync".

...Jim Thompson

G > Interesting problem. G > Kudos to mpm (alias Rube Goldberg) for the ball bearing solution. =A0LOL

G > Don't they still make LED's with blinking circuit built in? G > Do those have good common node rejection? G > Do they end up synching up with each other?

JT > I think it's a given that, unless you have phenomenal VDD filtering, JT > relaxation oscillators tend to at least "semi-sync".

Would the blinker LED's at least make the amount of filtering required more readily attainable?

G > Could the LED's with blinking circuit inside be used G > to drive FET driver transisters?

Jim Thompson seems to forget that he can't read my posts, so he hasn't actually read what I posted on the 19th. It's not a fully worked out solution - I haven't specified a clock generator to clock the pseudo- random binary sequence generator, nor specified the the CMOS shift registers that you might use to build it or the "n" and "m" numbers for a particular pseudo-random binary sequence generator - but I could do it in a couple of minutes. It isn't exactly rocket science, even if John Fields finds it intimidating.

-- Bill Sloman, Nijmegen

The one I posted on the 19th, which John Fields doesn't seem to like - a pseudo-randon binary sequence generator,

-- Bill Sloman, Nijmegen

ash

emi

The OP want 25 lamps. Twenty-five times five or six parts is 125 or

A pseudo-random binary sequence generator is a little more economical

- fewer parts and fewer solder joints.

-- Bill Sloman, Nijmegen

A 555 at each location allows local "stiff" filtering of the "Control Voltage" pin, which should help a lot to prevent syncing. ...Jim Thompson