I could use some help on this small circuit design - will it work?

[Adder_Noir]

Hi!

Please forgive me but this is the dodgiest circuit drawing ever made by anyone ever anywhere but I only had MS paint to hand

http://img514.imageshack.us/img514/3940/dodgyelectronicspic.png

It's a circuit which I want to use to make a small power circuit pulse at max voltage a variable number of times over a set time period. Basically instead of varying the voltage through the load I want to keep the voltage maxed but vary the time it is present for.

My idea is to use a variable resistor in an R-C arrangement to cause a capacitor to build up above the voltage required to overcome a zener diode and then activate a transistor input to allow current through the load briefly before the capacitor discharges enough to allow the the zener to stop it driving current again and then the cycle repeats.

I have 2 questions:

1) Is the circuit design good enough to allow the transistor to turn fully on for long enough or will the voltage constantly oscillate around the zener voltage causing the transistor to flutter on and off too quickly/weakly?

2) Is it possible to get a suitable power transistor to work in this fashion or am I asking too much of a transistor here? Will I have to combine it with an op amp or something to get it to work properly? The load in this case is a small model railway engine which takes 12 volts.

Thanks! ;)

 

[berkeman]

I don't think it will do what you want in its current form. If the charging voltage is constant, then the base drive zener curcuit will just stabilize at some value where the charging current equals the base current. I don't see a mechanism for oscillation in the current circuit.

What frequency and duty cycle do you want to achieve? You can use a simple CMOS (Schmidt trigger input) inverter and an RC to make an oscillator...

 

[Adder_Noir]

Thanks you make much sense as always. I can see the same thing clearly now you have pointed it out. It will as you say surely stablise and needs a mechanism to oscillate.

Do you have a link or something where I can find info on this CMOS set-up? I don't want to just be told the answer - I'll learn nothing and burden others!

I was thinking of working with a time period of around half a second and having a minimum of say 50 pulses up to a maximum of say a few hundred during that half second. How easy is it to vary the oscillation of a CMOS set-up. Can I use the variable resistor and capacitor arrangement to affect it's oscillation period?

Thanks.

 

[berkeman]

I mention the RC relaxation oscillator hookup using a Schmidt trigger inverter in this old thread:

https://www.physicsforums.com/showthread.php?t=87926

You can also look at 555 timers, which would probably be an okay building block for you. I've never liked them, but for your application, they may work okay.

 

[dlgoff]

You can also look at 555 timers,... I've never liked them, ...

Just curious. Why?

 

[Averagesupernova]

Just curious. Why?

I would be curious as well.

 

[Adder_Noir]

More good stuff, thank you berkeman

 

[Adder_Noir]

This is a brilliant link:

http://www.doctronics.co.uk/555.htm"

 

[berkeman]

I would be curious as well.

Mostly because they have such poor accuracy. Most of the oscillator functions that I need have to have pretty good accuracy. Not necessarily xtal accuracy for them all, but I don't like having the inaccuracy of the 555 trip points multiplying the tolerance of the R and C components. If I do a low-accuracy oscillator, it will generally be with a Schmidt trigger inverter or other simple oscillator.

It would be more useful if it didn't add in more inacuracy to the oscillation function, beyond the tolerance of the external R and C components, IMO.

 

[Adder_Noir]

Nice points berkeman cheers that may well come in handy!

This discussion is continued on the latest theoretical area it has wandered into https://www.physicsforums.com/showthread.php?t=374993"