Reducing current

Yes, you're missing the point. The world is full of alarm systems that likely have single-chip solutions to exactly what you need at a price cheaper than dirt. Get Bill Gates to fund it.

If you're determined to tilt the windmill... Put the speaker in series with a unijunction transistor. The time constant on the gate drive determines the frequency. Won't go bleep, bleep, but you can figger out how to live with a single audio tone.

Now, figger out how to synthesize the unijunction function from two transistors.

Now, figger out how to stop the oscillation with the least quiescent current drain.

Google some of the keywords above and you'll find lots of info.

• Bad. "common" does no mean "ALL NPN". Idea DOA.

• 1) use short pulses, 2) use piezo sounder.

• See above.

• BTW, >5years for batteries even if NOT used, is almost impossible due to tropic temperatures...

Looking at some of the back issues of magazines like Elektor would reveal such circuits. They classified the commonest transistors as TUP or TUN and used those terms in the schematics they published. They generally worked well with any of the devices from the long list of suitable devices.

inimal component count, and use only components that will be in common use for a long time, so no specific ICs. Also no specific trs, I want this to w ork with the majority of trs found in scrap. Hence I'm assuming a beta of 2

a bleep bleep. The base drive of one is shorted to ground with the wire lo op. To minimise current waste (it'll be battery powered) one side of each m ultivib will use a darlington. But its not getting me as low current as I w ant to go...

s 25mA base i needed. So the multivibrator passes collector i = 50mA. Bet a 25*25 means base i for that darlington is 0.08mA. This flows all the time when the alarm is guarding, and there are 2 isntances of this making 0.16m A quiescent current. That gives battery life of 6000hrs per Ah.

ast 5-10yrs untriggered would be a real plus.

e, and would require a long list of pinouts and are more likely to become o bsolete than discrete trs.

It looks like a darlington triple solves the problem. That raises one quest ion: how much current do I need to leak from b to e to keep jellybean trans istors switched off in tropical heat? I presume they'll be fine with 1M, bu t if I know how much leakage actually occurs I can probably lose 2 componen ts.

thanks

NT

Wouldn't it make a lot more sense to just tell them to use beta>=100 transistors selected for the critical positions?

(or to throw away any with beta < 40 except for the power stage)

Even the humble BC107 was rated for beta >100 @ 2mA.

Strikes me that a CMOS 555 or two would easily walk this job.

minimal component count, and use only components that will be in common us e for a long time, so no specific ICs. Also no specific trs, I want this to work with the majority of trs found in scrap. Hence I'm assuming a beta of 25 in real life conditions - good enough for a lot, but not all trs.

the less time, skill & equipment required, the more people can do it. In so me places there will be so little scrap available that selecting is not des irable.

under some conditions, sure.

ke a bleep bleep. The base drive of one is shorted to ground with the wire loop. To minimise current waste (it'll be battery powered) one side of each multivib will use a darlington. But its not getting me as low current as I want to go...

ans 25mA base i needed. So the multivibrator passes collector i = 50mA. B eta 25*25 means base i for that darlington is 0.08mA. This flows all the ti me when the alarm is guarding, and there are 2 isntances of this making 0.1

last 5-10yrs untriggered would be a real plus.

ive, and would require a long list of pinouts and are more likely to become obsolete than discrete trs.

rs.

Lots of ICs would, if I had free run I'd probably choose a hex invertor to do the lot. But its better to have something where the parts are easy to fi nd, can still be built in 30 years, and something that's easy to desolder a nd not static sensitive.

NT

I'm not seeing how that would enable people with nothing to generate income

not a good plan.

Interesting, ta

NT

thanks, I quite like that idea. The only issue with producing a TUN & TUP list is it would be very long, and over time an increasing number of trannies found in products would not be on it. I dont see a solution there.

NT

minimal component count, and use only components that will be in common us e for a long time, so no specific ICs. Also no specific trs, I want this to work with the majority of trs found in scrap. Hence I'm assuming a beta of 25 in real life conditions - good enough for a lot, but not all trs.

of course.

its working ok so far

ke a bleep bleep. The base drive of one is shorted to ground with the wire loop. To minimise current waste (it'll be battery powered) one side of each multivib will use a darlington. But its not getting me as low current as I want to go...

ans 25mA base i needed. So the multivibrator passes collector i = 50mA. B eta 25*25 means base i for that darlington is 0.08mA. This flows all the ti me when the alarm is guarding, and there are 2 isntances of this making 0.1

short pulses loses volume. A piezo is an option where little output is wanted, I expect that to be the minority of cases though. It needs to work with any of the main transducer types plus, homemade moving iron speakers.

last 5-10yrs untriggered would be a real plus.

transducer type makes no difference to battery life of an untriggered alarm .

If the battery lasts as long as it would on the shelf, I'm happy. I think t he triple should achieve that pretty much.

Thanks everyone

NT

use only components that will be in common use for a long time, so no specific ICs. Also no specific trs,

I want this to work with the majority of trs found in scrap. Hence I'm assuming a beta of 25 in real life

conditions - good enough for a lot, but not all trs.

snip

But its better to have something where the parts are easy to find, can still be built in 30 years, and

something that's easy to desolder and not static sensitive.

This is gonna sound like I'm being critical...because I am...in a good way...to be helpful. I've seen a lot of ideas come and go. People pick a tiny part of a solution to a problem that may or may not exist and head off, with the best of intentions, to make it so.

I'm all for helping the disadvantaged. I'd like to hear more about the big picture.

There's a thing known as a business plan. Before you jump up and start exclaiming that this is not a business... chill and listen.

The business plan describes the environment, the customer, the problem and how the business to be developed solves that problem in the CUSTOMER's context.

It discusses alternative solutions and why this one is superior and can succeed in the face of current practices and competitive solutions.

It describes the product, why the customer needs it, how he will learn about it and why he can/will pay what it costs.

It discusses the big picture of how this fits into the existing environment with all it's nuances.

It describes the manufacturing/distribution process. Labor force. Training. Quality control. Power source. It describes the suppliers, partners, synergies. How many will be built? Who's gonna install it? Who's gonna fix it when it breaks?

It has a development goal/schedule/development cost/product cost/price. Documentation? Budget? Schedule?

I'd question your constraints. Does it really make sense to restrict solutions to something that can be built 30 years from now? I'd suggest that you'll actually build very few of these.

The business plan is the why and the how of the project in relevant CONTEXT. It applies to every endeavor. You could apply the concepts to planning a marriage. It certainly would reduce the number of them.

Here's the result of a two-minute google search.

I disassembled one of these:

Has two transistors and an 8-pin chip under a blob of goop. Also has a relatively large inductor. Nothing is marked, but we can determine that it is possible to buy a similar alarm, with a magnet, and batteries, in quantity one shipped free from China for a buck. That suggests that, whatever it is under that blob of goop can be had for CHEAP!!!

Just because you may have access to obsolete electronics, doesn't mean that it's optimal to use 'em. I suggest it would be far easier to raise $1000 to buy 1000 finished devices, at retail, than to train people to build a thousand at a hundred different locations.

So, the business plan answers questions like:

What is the purpose of the project? Gainful employment? Is this the best use of their time? Are they awash in clean water? Medical care? Schools?

DIY project for the end user to produce something he needs? I've known degreed engineers who couldn't bias a transistor or solder two wires together.

Produce a product for sale to others. I've known degreed engineers who couldn't bias a transistor or solder two wires together. The customer's alternative is the device linked above.

Other? The answer to this question is CRITICAL. You must articulate the actual objective.

Who's gonna answer the alarm and what can they do about neutralizing the threat?

Is the need real, or just someone's view of PART of a much bigger need. Building half a bridge doesn't help traffic get across the water.

Given the availability of raw material, skills of the workers, need for the result, ability to pay for the result (cash barter trade other)...is this really the best use of everyone's time?

I admit that I have no direct experience with the horrific conditions that exist in much of the world. I expect that anyone reading this can afford a computer and would not last a week in that environment. When we try to help, we need to give them what they need, not what we envision as a fun project.

One of the widespread problems in the world of "helpers" is tunnel vision. Need an alarm. Let's break a wire and use some transistors.

Ok, what if we take a more simplistic approach? Let's make the current zero.

Wind up a stick with a trip wire that lets the stick strike the perp. His scream is the alarm.

Rock on a string wrapped around a stick that, when released causes a smaller rock to rotate and flail on something that makes noise?

Something like an animal trap.

Container of water that gets uncorked by the trip wire and flows in the face of the person responsible for watching the place?

Solutions that fit familiar situations are more likely to succeed.

One constant that's often ignored in the security industry... You don't have to prevent a theft. All you need to do is make your place look less attractive than your neighbor's.

Bottom line is that I don't think you have an electronic design problem...but you have not disclosed enough about the big picture to tell for sure.

The weakest link in any project is the PLAN.

Make toe poppers. Very effective alarm.

You mean well. The short answer is yes we have looked at these questions, and yes we do want to make it possible to make an assortment of electronic goods for nothing from scrap. It serves a few purposes:

- makes the person enough income to survive

- teaches them elec eng, which if they gain enough ability can get them out of poverty for good

- provides alarms that improve life a little, and does so for less cost than those chinese things

NT

possible to make an assortment of electronic goods for nothing from scrap. It serves a few purposes:

Ok, but if you have crunched the numbers, expressing them would be helpful as a model for other altruists.

It would be interesting to know at least some of the relevant numbers. How much income would one receive from building one of these? How many hours would it take to build one? How would they be marketed? Who would buy it and where would they get the money? Of course, a relevant point is the total income of one such person in that economy. I'm still concerned that transporting a hundred pounds of scrap electronics so you can pull out a six-cent transistor then get rid of the rest might end up costing more than just buying a built device.

How many did you plan to build? That's a critical assumption.

I think we may have a different definition of elec eng. Who's gonna teach it? If my experience with college is any indication, very few of those "teachers" can teach. I learned more my first week on the job than in 5 years of college. You propose to do it all with a "project".

Just reading any newsgroup should convince you that many very smart people can't compose a coherent thought, much less teach electronics to the uneducated with a completely different language and culture REMOTELY.

I interact with some local hobbyists. Most of their questions could be answered by a simple application of Ohm's Law. And I've walked them thru the process dozens of times in the past. Teaching is HARD. You can't just accost anybody on the street and make them an engineer.

If you've ever constructed a Heathkit, you know how much documentation is required when you supply all the parts. Imagine a Heathkit with no parts and sparse generic instructions to get some random part, figger out the pinout and try to hook it up like this schematic. I submit that even finding acceptable resistor values will be problematic. If you don't have a mentor on site, you're in big trouble.

I applaud your goals. I think your expectations exceed the resources available. You've resisted requests to provide sufficient details to challenge that assessment.

Of course, you don't require my blessing to proceed. I was a project engineer for a long time. I've watched many people attempt to do the wrong thing for the right reason. I'm just trying to help.

s, and yes we do want to make it

out of poverty for good

than those chinese things

on.

These were good questions a year and a half ago.

When young I taught myself enough elec eng from one book to build simple ci rcuits, repair things with a patchy but useful level of sucess, and make en ough per hour to rise above the World Bank's definition of poverty. So did lots of people on this ng. No mentor, no internet, no newsgroup, nothing. S ome people can do that.

I've got everything working on the alarm with 7 transistors, except a low b attery warning, that'll make 9. I had hoped to use less, but it seems that' s what it takes if you want decent volume and battery life.

I'll probably do an even more cut down 3 tr version with a piezo too. Thank s everyone.

NT

Not speaking from a 'knowledgeable' position, but more of a 'thought process' position...Isn't MOST of the electrical energy expended in a speaker spent in the coil, NOT launching energy into the air? Which implies using some type of resonance to increase the current without increasing the current consumption, so to speak. As I said, I don't know a lot about this electrical to acoustic energy transducers.

The other thought, years ago I was asked to improve a 'dead man sounder' which is battery device clippled to a fireman, if he goes down and doesn't move for awhile, this thing goes off nnd 'squeals' at over 150dB [maybe only 125dB, but really, really loud!] The concept came about after firemen lost a colleague fireman who had on oxygen equipment, went down unconscious, and could NOT be found in the smoke filled room and muck/debris. They could NOT find him in this small room over a very long period of time; his oxygen depleted and he tragically died. Thus, came about the concept of this 'dead man' device.

What I learned was that there is great magic in acoustical design. The amplifying characteristics, or rather improving the efficiency at energy conversion, was impressive. Very small diaphragm, VERY LARGE NOISE! Try to do a search for the manufacturers of thses types of devices, [of course their batteries are usually fresh and their use is maybe once, but there might be some real insights in that electrical/acoustical interface] After all, your alarm sound is NOT broadband, so why use a broadband device? Even old car horns consisted of nothing more than a metal diaphragm attracted to a coil, and when the diaphragm just starts to touch, or is that touch?, the current is mechanically switched OFF and the diaphragm springs back to position only to have the current come back ON and cycle start again. I know it's a mechanical solution, but you said you wanted 'available' technology, can't get much more available than that.

arrrggg! backwards as usual...I think the old horns sat with their diaphragms agains the connection, ready to put current through the coil. when you hit the horn ring on your steering wheel, current flowed through the horn and the coil PULLED the diaphragm away from its contact, thus circuit was broken and the diaphragm no longer being pulled away, would snap back and make contact keeping the cycle going.

I do commend the idea of using scrap pcb's, and I a have a couple boxes of them, they usually are only used when I don't have what I need in the parts I have. And it most often is a capacitor that I pull off the board, transistors are just so cheap. You can buy 10,000 2n3904 transistors for $160 or 1.6 cents each. How much electricity, solder suckers, wicking material, scrap pcb hauling, soldering iron wear and tear can you have for 1.6 cents?

Mikek

Re speakers, moving coils are only a few percent efficient - the smaller, the less efficient. That's why I mentioned home made sounder options, meaning moving iron, which are greatly more efficient, and can be made with more or less no inbuilt damping.

The horn idea is interesting. It could possibly drop component count to 2 transistors. Only downside is I cant spice it, will have to make one! Cheers for that.

NT

You can :) Not going to happen when you're struggling to find food by the roadside.

no electricity, no sucker, no wick, no olsdering iron in the sense that we know them will be used. And scrap nails it doesnt matter if they wear away after so many years.

NT