Single transistor battery disconnect?

Hi All,

This is my first post so a quick Hello to you all and a big thanks for taking the time to read this..

Quick Summary of my question

I wish to modify the attached circuit so the relay activates and deactivates at two different voltages. Ideally using just a few extra passive components. I am sure it is possible but need some help finding a solution...Ive thought about this for a while and been unable to find a good solution. (I know there are many ways of achieving the desired end but I have to use a transistor and handful of components - nothing fancier, I could buy an IC to do the job but that is not an option)

The pdf shows the basic idea and I believe this circuit will energise the relay coil when the power supply is at approx the same voltage as the zener diode breakdown voltage.

But how can I make the voltage at which the coil de-energises a few volts lower? Could I use the second set of relay contacts to switch in a lower voltage zener diode? There must be another way.

OriginallyI tested a circuit without the transistor. Just a zener diode in series with a relay. This did achieves the desired effect. But power dissapation in the zener will be very large if the relay current is high. And I am concerned that if it were manufactured in quantity the on/off voltages may change from one unit to the next due to fluctuations caused by component tolerances and environmental heating/cooling effects.

Please if you have any ideas/comments I am very keen to hear them? Any advice is very welcome.

Regards
PhilVr

What is the goal?

Some one could test a circuit or simulate it if given the pull-in, drop-out voltage of your relay. The coil specs. Also the desired pull-in, drop-out voltage.

A relay I just tested by itself was 14V pull-in and 4V drop-out.

Are you looking for a small difference?
In other words, Schmitt Trigger function, which gives a wide range. Or a sharp threshold level?

ps.
A single transistor?

I am looking for a difference of about 2 or 3 volts - but i hope once the circuit is designed it will be trivial to change the voltage range. I have had some ideas (expressed below) but the more I think of it the more concerned I am at the prospect of a see-saw oscillation as the power dis-connects and re-connects ....I need to press on and prove the basic idea below works and then worry about the oscillation...Ta for your comments Inwo!


As mentioned by Inwo above, relays have different drop-in and drop-out voltages. I fear these may vary between relays and also with temperature and zener diodes are often quoted as having a 5% tolerance. I dismissed the most basic circuit for these reasons and also because the diodes would dissipate a lot of power if the relay coil required a large current - a 12V coil with 200mA current is nearly disapating 2.5W, an 18V coil is 3.6W so a 5W zener will get hot - the next sizes of zener are 10W and 20W and they are rather difficult to obtain quickly and they are big!

The goal is to create a circuit that will disconnect a voltage (power) source when its voltage falls below a preset threshold and automatically reconnect it when the voltage source increases above another voltage threshold.

Power source could be a battery or a bank of super capacitors or any source that needs protection from being drained below a certain voltage.

I know there are many ways this can be achieved and no doubt with greater ease. But the solution I seek has to use through hole technology and be able to fit within an an existing pcb.

I have seen a device by Linear Technology that does exactly what I want the "ltc2960" and would use it if I was able to design a new pcb ( see: http://www.linear.com/product/LTC2960).

So I am looking for the simplest solution possible using a minimal quantity of through hole components. I attach a second circuit idea - comments very welcome...I will try simulating when I gain access to a suitable computer (with LT-Spice) and report back.

With reference to relay-circuit-1.pdf which I attach...

The two zener diodes and a pnp transistor are arranged so the pnp transistor will become active when the coil is energised and bypass one of the zeners thus (hopefully) lowering the voltage at which the relay coil is switched off...(untried! I dont know if it works yet - just an idea). When the voltage increases the pnp is inactive so the switch on voltage is aprox the sum of the two zener voltages

You might now have guessed that I am not too hot on designing with discrete components. In that case you are correct. I think that discrete components are interesting and could be quite a lot of fun to work with...I have so much to learn.

thanks in advance for reading this..
kind regards
Philvr

Now that you've opened it up to two transistors.
Google transistor schmitt trigger relay.




https://www.google.com/search?q=tra...onics.com%2Fpay%2FBEC-2%2FPage32.html;285;240